Seton Hall UniversityeRepository @ Seton HallSeton Hall University Dissertations and Theses(ETDs) Seton Hall University Dissertations and Theses
2010
Cost and Price Increases in Higher Education:Evidence of a Cost Disease on Higher EducationCosts and Tuition Prices and the Implications forHighes Education PolicyJerry TrombellaSeton Hall University
Follow this and additional works at: https://scholarship.shu.edu/dissertations
Part of the Economics Commons, and the Higher Education Commons
Recommended CitationTrombella, Jerry, "Cost and Price Increases in Higher Education: Evidence of a Cost Disease on Higher Education Costs and TuitionPrices and the Implications for Highes Education Policy" (2010). Seton Hall University Dissertations and Theses (ETDs). 351.https://scholarship.shu.edu/dissertations/351
Higher Education and the Cost Disease
Cost and Price Increases in Higher Education: Evidence of a Cost Disease on Higher Education Costs and Tuition Prices and the
Implications for Higher Education Policy
by
Jerry Trombella
Dissertation Committee
Dr. Joseph Stetar, Mentor Dr. Beth Castiglia
Dr. Rong Chen Dr. Martin Finkelstein
Submitted in Partial Satisfaction for the Requirements for the Degree of Doctor of Philosophy
Seton Hall University
2010
Higher Education and the Cost Disease
O Copyright by Jerry Trombella, 2010
All Rights Reserved
SETON HALL UNIVERSITY COLLEGE OF EDUCATION AND HUMAN SERVICES
OFFICE OF GRADUATE STUDIES
APPROVAL FOR SUCCESSFUL DEFENSE
Doctoral Candidate, Jerry Trombella, has successfully defended and made the required
modifications to the text of the doctoral dissertation for the Ph.D. during this Summer
Semester 2010.
DISSERTATION COMMITTEE @lease sign and date beside your name)
Mentor: Dr. Joseph Stetar r 7// 4 4
Committee Member: Y Dr. Martin Finkelstein
Committee Member: Dr. Rone Chen
Committee Member: Dr. Beth Castiglia
The mentor and any other committee members who wish to review revisions will sign and date this document only when revisions have been completed. Please return this form to the Office of Graduate Studies, where it will be placed in the candidate's file and submit a copy with your final dissertation to be bound as page number two.
Higher Education and the Cost Disease i
ABSTRACT
As concern over rapidly rising college costs and tuition sticker prices have
increased, a variety of research has been conducted to determine potential causes. Most
of this research has focused on factors unique to higher education. In contrast, cost
disease theory attempts to create a comparative context to explain cost increases in higher
education. The theory postulates that all heavily labor-intensive industries will experience
faster than average cost increases, based on the limitations in leveraging technology to
increase productivity. This research attempts to analyze the extent to which a cost disease
affects college costs and tuition sticker prices in two distinct segments. First, trend
analysis is used to analyze components of the higher education price index from 1961
through 2008 to assess the extent to which labor costs have driven higher education costs
over time. Second, changes in higher education costs and tuition prices are compared
against components of the Personal Consumption Index of the National Income and
Product Accounts from 1961 through 2008 to determine the extent to which a cost
disease differentially impacts labor intensive sectors of the economy.
. . Higher Education and the Cost Disease 11
AKNOWLEGEMENT
I would like to acknowledge the members of my dissertation committee for all
their assistance in completing this dissertation. Dr. Stetar helped me take an initial
inchoate topic and guide it toward completion. Dr. Finkelstein provided key insights and
allowed me to explore initial ideas in an earlier class. This became the basis for the
dissertation topic. Dr. Castiglia and Dr. Chen provided critical feedback which helped
create a cleaner, crisper final product. Thank you again for all your encouragement and
support.
I would like to dedicate thls dissertation to my parents. Although their own
education was interrupted by poverty and war, as immigrants to the United States they
nevertheless instilled in their children the importance of an education. Thank you for all
your sacrifice and support.
Higher Education and the Cost Disease iii
TABLE OF CONTENTS
ABSTRACT ......................................................................................................................... i . . AKNOWLEGEMENT ....................................................................................................... 1 1 ... TABLE OF CONTENTS ................................................................................................... 111
TABLE OF TABLES .................................................................................................... vi CHAPTER I: INTRODUCTION ........................................................................................ 1
Statement of the Problem ................................................................................................ 2 Research Questions ......................................................................................................... 6 Importance of the Topic ................................................................................................ 11
CHAPTER 11: REVIEW OF THE RELATED LITERATURE ....................................... 16 Factors Unique to Higher Education ............................................................................. 16
University Goals. Pursuit of Excellence and the Revenue Theory of Costs: Valuation and Higher Education Expenditures ........................................................ 16 University Goals and Competition for Excellence .................................................... 19 Wealth. Market Structure. and Competition ............................................................. 21 Merit Aid. Tuition Discounting. and Rising Costs .................................................... 24
Inherent Production Cost Factors ................................................................................. 28 Cost Disease of the Service Sector ............................................................................... 30 Literature Review - Prior Methodology ....................................................................... 40
CHAPTER LII: METHODS ............................................................................................. 44 Conceptual Framework ................................................................................................. 44 Research Methods ......................................................................................................... 48 Data Sources .......................................................................................................... 49
Role and Uses of Price Indexes ................................................................................ 49 Personal Consumption Expenditure by Product Categoiy ....................................... 50 Higher Education Price Index .................................................................................. 52 . .
Statist~cal Methods ........................................................................................................ 54 Significance and Limitations of the Study .................................................................... 58
Signzficance of Study ................................................................................................. 58 Limitations of the Study ............................................................................................ 59
CHAPTER IV: RESEARCH FINDINGS ........................................................................ 62 Higher Education Price Index: Integrating HEPI Data Series ...................................... 62 Research Question 1: What are the main cost drivers responsible for driving the Higher Education Price Index? ................................................................................................. 69 Research Question 2: To what extent are labor costs driving overall costs within higher education? ..................................................................................................................... 80 Research Question 3: To what extent can a cost disease explain rapidly rising costs and . . tuit~on sticker prices? .................................................................................................... 82
Price increases all PCE goods and services =price increases PCE durable goods ..................................................................................................................... 89 Price Increases all PCE goods and services =price increases PCE non-durable goods ............................................................................................................... 90
Higher Education and the Cost Disease iv
price increases all PCE goods and services =price increases PCE services ..... 90 price increases all PCE goods and services = higher education price increases.91 price increases all PCE goods and services = higher education cost increases .. 91 price increases PCE durable goods =price increases PCE non-durable goods . 91 mice increases PCE durable poods =price increases PCE services .................. 92 . price increases PCE durable goods = higher education price increases ............ 92 price increases PCE durablegoods = hipher education cost increases .............. 93 . . price increases PCE non-durable goods =price increases PCE services ........... 93 price increases PCE non-durable goods = higher education price increases ..... 93 price increases PCE non-durable goods = higher education cost increases ....... 94 price increases PCE services = higher education price increases ....................... 94 price increases PCE services = higher education cost increases ........................ 95 higher education cost increases = higher education price increases ................... 95
Research Question 4: Are there similarities between increases in higher education costs and tuition sticker prices and prices in other labor intensive industries? ..................... 97 Research Question 5: Are there differences between price increases in labor intensive industries compared to those associated with the manufacturing sector? ..................... 99
CHAPTER V: CONCLUSION ....................................................................................... 102 The Cost Disease and Higher Education: Evidence from Research Findings ............ 102 Theoretical Implications of the Research ................................................................... 105 Increases in Higher Education Prices relative to Costs .............................................. 107
Tuition Discounting ............................................................................................ 111 Discussion and Analysis ............................................................................................. 113 Public Funding: Finding the Balance between Private Benefits and Public Positive . . External~ties: ............................................................................................................... 116 Policy Recommendations ............................................................................................ 118
Fundingfor Higher Education Need-Based Aid Programs ................................... 118 Institutional Tuition Discounting and Public Policy ............................................. 120 Increasing Productivity .................................................................................... 121 Increasing Productivity: Technology and On-Line Classes. Possibilities and Limitations .............................................................................................................. 122
Increasing Efficiency .................................................................................................. 126 Increase High School Pro$ciency .......................................................................... 128 College Credits in High School ........................................................................... 129 Three Year Undergraduate Degree ........................................................................ 130 Financing of Higher Education .............................................................................. 131 Allocation of Higher Education Public Funding: Institutional Subsidies. and State Need and Merit-based Student Assistance Grants .................................................. 132 Institutional Subsidies vs . Student Need-based Grants ........................................... 134 Income Contingent Loans ....................................................................................... 135 Administrative Salaries and the Costs of Regulation ............................................. 140
Limitations of the Study and Areas for Future Research ............................................ 141 Conclusion .................................................................................................................. 145 . . Appendix A: Defimtlons* ........................................................................................... 159 Appendix B: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment ............................................................................................. 161
Higher Education and the Cost Disease v
Appendix C: Consumer Price Index, Higher Education Price Index, and Major ........................................................................................ Subcomponents, 1961-2001 168
Appendix D: Consumer Price Index, Higher Education Price Index, and Major .............................................. Subcomponents, 1961 -2001, Reindexed, 1961 = 100 175
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index, 1929 - 2008 ............................................................................... 182
.... Appendix F: National Income and Product Accounts, Reinexed, Base Year 1961 215 ...................... Appendix G: Research Database Used to Conduct ANOVA Analysis 236
Appendix H: Future Research: Creating a Model of Cost and Price Escalation in ........................................................................................................ Higher Education 238
Higher Education and the Cost Disease vi
TABLE OF TABLES
Table 1. Regression Model Used to Calculate Contracted Services, Supplies and Equipment. .......................................................................................... .65 Table 2. Percent Distribution of College and University Current Fund Educational and General Expenditures, Budget FY 1983.. ........................................................ 68 Table 3. Historical Summary of the Consumer Price Index and Higher Education Price Index. ................................................................................................ 69 Table 4. Consumer Price Index, Higher Education Price Index, and Major HEPI
...................................................................... Subcomponents, 1983-2008.. 71 Table 5. CPI, HEPI, Professional and Non Professional Salaries, Fringe Benefits, Total Personal Compensation and Total Contracted Services, Supplies and Equipment.. ....... 73 Table 6. Consumer Price Index, Higher Education Price Index, and Major HEPI
..................................................................... Subcomponents, 1961 -2008.. 77 Table 7. Consumer Price Index, Total Personal Compensation and Contracted Services, 1961 -2008.. .......................................................................................... 8 1 Table 8. ANOVA Analysis: All PCE Goods and Services, Durable Goods, Non-Durable Goods, Services, Higher Education Prices, and Higher Education Costs.. ................ .86 Table 9. Results of Post Hoc Multiple Comparison Test.. .................................... 87 Table 10. Percent of All First-time Full-time Entering Students at 4-Year Institutions Completing Undergraduate Degrees: Fall 1996 Through Fall 2001 Cohorts.. ......... ..I26 Table 11. 150 Percent Certificate or Associate Degree Completion Rate for Full-time Degree Seeking Students: 1994 Through 2004 Starting Cohorts.. ........ ,127 Cohorts
Higher Education and the Cost Disease 1
CHAPTER I: INTRODUCTION
Over the last 45 years, college tuition prices have been rising approximately twice
as fast as the rate of inflation, as measured by the consumer price index. Students, parents
and members of Congress have been expressing growing unease about the affordability
of a college education, and rising frustration with colleges over tuition increases. Over
the last decade, Congress has mandated a variety of studies to determine the causes for
these faster than inflation increases, and has increasingly threatened regulatory solutions
in an attempt to curb what is seen as run-away tuition prices.
As concern over tuition prices has increased, a great deal of research has been
conducted to determine potential causes. Much of the discussion has centered on
identifying factors unique to the higher education industry which may be responsible for
excessive price increases. A great deal of this research has focused on the extent to which
the non-profit structure of the vast majority of higher education institutions has
influenced institutional goals and aspirations, while fueling needless spending and
engendering excessive competition among institutions.
In contrast, a parallel body of research has focused on the extent to which rapidly
rising tuition prices are related to a much broader phenomenon potentially affecting all
labor intensive industries. Known as Cost Disease Theory, this line of inquiry postulates
that all labor-intensive industries will experience faster than average cost increases, based
on limitations in leveraging technology to increase productivity.
Determining the causes for rapidly rising tuition prices may have important policy
implications; Federal and state public policy and resource allocation choices affecting
higher education will hinge on the assessment of the causes for tuition increases. If
Higher Education and the Cost Disease 2
escalating costs are caused by institutional greed or a skewed priorities leading to
excessive spending, then Congress may justifiably impose regulatory solutions as a way
to control institutional spending, or may shift resources away from higher education to
more pressing public policy issues. However, if higher tuition is due to factors beyond
institutional control, as described by a cost disease, such causes, if properly understood,
may actually justify an enhanced public reinvestment in institutions of higher education.
This research examines the competing theories attempting to explain cost and
price increases in higher education, and attempt to determine the extent to which a cost
disease can account for cost increases in higher education, and by extension, higher than
average tuition price increases. If a cost disease is primarily responsible for driving
higher education costs, evidence for this should be found by deconstructing components
of the higher education price index, as well as by comparing higher education price
increases to those in other industries. If a cost disease theory is correct, faster-than-
inflation price increases should also occur in other labor intensive industries, irrespective
of whether the firms in these industries are predominantly non-profit entities or profit
oriented firms.
Statement of the Problem
Concern over rapidly rising tuition prices is not new; just 1 .year after passage of
the Higher Education Act ("The Higher Education Act of 1965," 1965), Baurnol and
Bowen (1966) were already speculating on the potential causes for rapidly rising tuition
prices. By 1973, the Carnegie Commission on Higher Education reported that, between
1929 and 1960, the cost per credit hour at colleges and universities increased an average
rate of over 2 %% above the consumer price index (1978). Congress created the National
Higher Education and the Cost Disease 3
Commission on the Financing of Postsecondary Education in 1972, in part because of
Congressional concern over rapidly rising college costs (Finn, 1978). By 1990, a
Brookings Institution study noted that between 1980 and 1987, tuition prices rose at twice
the inflation rate, which was lo%, even outpacing rapidly rising health care costs, which
grew at an average annual rate of 8% (Hauptman, 1990).
Anxiety over the rising cost of a college education, especially with the "sticker
prices" of published tuition and fees, became widespread during the 1990's, in part due to
the slow growth of family income relative to increases in tuition prices (Ehrenberg,
2000). Between 1981 and 2000, after adjusting for inflation as measured by the consumer
price index (CPI), tuition more than doubled at public not-for profit 4-year colleges and
universities, while median family income grew by 27% and financial aid per full-time
equivalent student increased by 82% (Horn, Wei, Berker, & Carroll, 2002). Students and
their families have been expressing growing unease about the affordability of a college
education; the proportion of adults expressing concern that qualified and motivated
students will not have the opportunity to receive a college education increased from 45%
in 1998 to a record high of 62% in 2007, and more than 75% of parents indicated anxiety
over paying for their children's college education (Immenvahl & Johnson, 2007).
Public concern about rising tuition prices led Congress to establish the National
Commission on the Costs of Higher Education in 1997, with a mandate to review college
costs and prices (Jones, 2001). As part of the 1998 Amendments to the Higher Education
Act, Congress mandated the study of expenditures at higher education institutions ("1998
Amendments to the Higher Education Act of 1965," 1998). Concern over the rate of
tuition sticker price increases became an increasingly important issue during successive
Higher Education and the Cost Disease 4
reauthorization hearings. With the average price of a college education increasing at
twice the rate of inflation between 1981 and 2000, some of the proposed legislation in
anticipation of the 2004 Reauthorization of the Higher Education Act even contained
provisions which threatened the loss of federal student assistance grants to colleges which
increased tuition more than double the inflation rate ("Affordability in Higher Education
Act," 2003). This was later amended to become a provision requiring higher education
institutions to report annual tuition and fee increases against a suggested "college
affordability index," defined as twice the rate of inflation. Institutions classified as
"excessive" would be required to provide reports, plans, and a schedule for controlling
future tuition charges ("College Access and Opportunity Act," 2003).
While the provisions failed to pass, they suggest the degree of concem over price
increases in higher education, articulated by both Congress and the public, and
uncertainty about the cause. Meanwhile, although much attention has been focused on the
issue of college cost and prices, intensified by Congressional and state mandated studies,
there is still a great deal of confusion concerning the nature of these faster-than-inflation
cost and price increases.
The concem over tuition increases reflects the unique role of higher education as
one of the most important mechanisms of equality of opportunity, and the anxiety reflects
public fear that a college education may be priced beyond the range of students and their
families. At its best, the attention on cost and price has stimulated an important
discussion about the function of higher education and its relationship to societal needs
and aspirations, along with an increased appreciation of the nexus between access to
higher education and equality of opportunity.
Higher Education and the Cost Disease 5
Unfortunately, much of the debate has been polemical, stemming from a belief
that colleges and universities have been profligate in their spending, and indifferent to the
needs of students and their families, as well as society at large. This feeling of frustration
and confusion was expressed in a recent report sponsored by Congressmen John Boehner
and Howard McKean, which stated, "Higher education is deemed such an essential piece
of the success puzzle, colleges feel justified in routinely kicking middle-America in the
teeth" by increasing tuition prices (Wolanin, 2005, p. 46).
Spurred by the attempt to curb tuition price increases, a considerable body of
research has been conducted to determine the most likely causes. Two broad sets of
causes have been offered to explain the faster-than-inflation costs associated with higher
education: (a) causal forces exclusively associated with higher education, and (b) the
Theory of the Cost Disease, which attempts to place increasing higher education cost, and
by extension, tuition price increases, within the framework of cost pressures facing all
labor-intensive industries.
In seeking to find answers to rapidly rising tuition sticker prices, much of the
focus has centered on whether there are factors unique to higher education which drive
cost increases faster than the consumer price index, which then become reflected in
higher tuition prices. While these studies have identified a number of potential "cost
drivers," which include the goals associated with higher education institutions,
accelerating competition among institutions, the growth of institutional aid, as well as the
increasing costs of technology, state and federal regulations, faculty compensation
workload policies, these are perhaps as unhelpful as they are comprehensive (Brennan,
2001, p. 15; Cunningham, Wellman, Clinedinst, Meriosotis, & Carroll, 2001, p. 21).
Higher Education and the Cost Disease 6
In contrast to explanations centered on factors unique to the higher education
industry, the Theory of the Cost Disease attempts to create a comparative context to
explain cost increases in higher education; the theory postulates that all heavily labor-
intensive industries will experience faster-than-average cost increases, based on
limitations in leveragmg technology to increase productivity. This provides a basis to test
the validity of the theory both in potentially explaining higher education cost pressures as
well as the extent to which a cost disease influences other labor-intensive industries
(Archibald & Feldman, 2008).
Research Questions
This research approaches assessing the extent to which a cost disease affects cost
and price increases within higher education in two distinct segments; cost disease theory
postulates that labor-intensive industries should experience faster-than-inflation price
increases compared to industries associated with the manufacturing sector. According to
the theory, this is directly related to the underlying nature of an industry's production
function. Unlike the manufacturing sector, where labor is incidental to the production
process, labor is itself the primary output associated with labor-intensive industries, and
therefore, less amenable to productivity increases associated with the introduction of
labor-saving technology. Researchers should be able to discern both cost and price
increases first, by examining relative changes in prices over time among the components
associated with the production process, and secondly, by changes in relative prices
among the industry under study compared more generally to those associated with other
labor intensive industries as well as those associated with the manufacturing sector.
Higher Education and the Cost Disease 7
This research utilizes two distinct datasets and two different analytical methods to
analyze the extent to which the presence of a cost disease can explain both cost and price
increases in higher education. First, trend analysis is used to analyze components of the
higher education price index to assess the extent to which labor costs have driven higher
education costs over time.
While evidence that the labor component of the higher education production
function has been driving higher education costs may provide important evidence
supporting the presence of a cost disease in higher education, by itself this is insufficient
evidence for a cost disease. The theory of a cost disease also implies that increases in
higher education costs should also reasonably mirror cost increases in other labor-
intensive industries, which should also be higher than those associated with industries in
the manufacturing sector.
Second, to examine the extent to which higher education is affected by a cost
disease, the researcher compares changes in higher education cost and prices to other
labor intensive industries, in addition to the manufacturing sector of the economy. This
provides some additional challenges. The Higher Education Price Index was specifically
created to analyze costs associated with components of higher education's production
function. Nothing quite like it exists among other consumer product categories. However,
there is a way to compare relative changes in prices in higher education to other
consumer purchases: the Personal Consumption Expenditures (PCE) Indices of the
National Income and Product Accounts, first created during the great depression as a way
to assist the Roosevelt Administration in developing and monitoring New Deal economic
policy, provides indices measuring the type of goods and services purchased over time.
Higher Education and the Cost Disease 8
These can be used to compare relative changes in prices of higher education compared to
consumer purchases associated with other service sectors, as well as those associated
with the manufacturing sector.
However, beyond reflecting intrinsic production costs, higher education prices
also encompass variability associated with relative changes in sources of support,
especially from changes in revenue associated with state subsidies for public institutions,
and gifl and endowment income for private institutions. Declining revenue streams
associated with public and private giving may exacerbate price increases during periods
when federal and state budgets are constrained, or economic circumstances or changes in
financial markets constrain private giving.
Thus, in attempting to assess the relative changes in prices across labor-intensive
and manufacturing industries (using the NIPA PCE data), the researcher also compares
higher education costs as well (based on HEPI data) using ANOVA. Including both
higher education cost and price indices when conducting relative assessments of a cost
disease among various sectors of the economy will help account for potential issues
associated with burden shifting.
The primary research question of this study is: To what extent can cost and tuition
sticker price increases in higher education be explained by the presence of a cost disease
affecting colleges and universities? Auxiliary questions include:
1. What are the main cost drivers responsible for driving the Higher Education
Price Index? This research utilizes trend analysis examining components of the higher
education price index over time. Analysis of the HEPI index reveals that higher education
costs significantly outpace price increases associated with the consumer price index.
Higher Education and the Cost Disease 9
However, among components of the higher education price index, certain sub-indices
may be increasing at a rate even faster than the HEPI average; Cost drivers are defined as
those components of the higher education price index which are increasing at a rate faster
than the aggregate HEPI average. This specifically addresses issues associated with the
cost of producing higher education and the relative changes in component categories over
time.
2. To what extent are labor costs driving overall costs within higher education?
This research utilizes trend analysis examining components of the higher education price
index over time. If labor costs are increasing faster than the aggregate HEPI index, they
will be considered cost drivers propelling higher education costs.
3. To what extent can a cost disease explain rapidly rising higher education costs
and tuition sticker prices? Research analyzing the relationship between a cost disease and
higher education cost and price increases involves two distinct components; first, using
trend analysis associated with auxiliary research questions 1 and 2, the extent to which
salary-related costs are driving total production costs within the higher education industry
are examined. Secondly, using ANOVA, cost increases associated with higher education
(as measured by the higher education price index), price increases associated with higher
education (as measured by the National Income and Product Accounts Personal
Consumption Expenditure Index) and price increases associated with durable goods, non-
durable goods, and services are compared to determine the extent to which both cost and
price increases mirror those associated with the other Service sector consumption items.
Cost Disease Theory suggests that price increases associated with service sector
industries should outpace those associated with the manufacturing sector, (represented by
Higher Education and the Cost Disease 10
NIPA PCE durable goods purchases), while non-durable goods, which are generally more
labor-intensive than those associated with Durable Goods, but less dependent than
services on labor should also have price increases higher than those associated with the
pure service sector. Additionally, Cost Disease Theory suggests that cost increases
associated with the Higher Education Price Index should outpace price increases
associated with Durable and Non-Durable Goods, while they should be similar to price
increases associated with Service Sector goods. ANOVA post-hoc multiple comparison
tests will be used to assess the extent to which statistical differences exist among higher
education HEPI cost increases, NIPA higher education price increases, NIPA Durable
good increases, NIPA Non-Durable Good price increases, and increases associated with
NIPA-classified service items.
ANOVA post-hoc multiple comparison tests in which HEPI cost increases, NIPA
higher education price increases, and NIPA Service Good increases are statistically
different (at .05 level of significance) from price increases associated with Durable and
Non-Durable Goods will be considered evidence that a cost disease impacts both higher
education in particular and service sector purchases more generally.
4. Are there similarities between increases in higher education costs and tuition
sticker prices and prices in other labor intensive industries? ANOVA is conducted to
assess the extent to which the NIPA Higher Education Personal Consumption
Expenditure Index mirror those associated with other Service Industries during the period
between 1961 and 2008. Cost Disease Theory suggests that Price Increases associated
with higher education (based on the NIPA Personal Consumption Expenditures) should
approximate those associated with the broader Services category of the NIPA Personal
Higher Education and the Cost Disease 11
Consumption expenditures. An ANOVA post-hoc comparison test will be performed for
the NIPA higher education PCE and the NIPA Service category PCE; price increases in
the higher education NIPA index and those associated with the broader NIPA Services
category will be considered similar if the ANOVA post-hoc comparison test (at a .05
level of significance) is not found to be statistically significant.
5. Are there differences between price increases in labor intensive industries
compared to those associated with the manufacturing sector? ANOVA is used to compare
both cost increases associated with the Higher Education Price Index, as well as price
increases associated with the National Income and Product Accounts for Higher
Education, Durable Goods, Non-Durable Goods and Services. The extent to which there
are similarities or differences between prices associated with durable and non-durable
goods and those of the service sector will provide basic evidence about the presence of a
systematic cost disease associated with service sector industries. An ANOVA post-hoc
comparison test will be conducted among the NIPA PCE price indices associated with
Durable Goods, Non-Durable Goods and purchases associated with the NIPA Service
category. After conducting an ANOVA post-hoc comparison test (conducted at the .05
level of significance), price increases in the manufacturing sector will be considered
different from those in the service sector if price increases associated with the NIPA
Service category are found to be statistically significantly different from those associated
with the NIPA Durable and Non-Durable Goods category.
Importance of the Topic
A variety of very different explanations have been offered for the causes of the
persistent price increases facing students and families, with important implications for
Higher Education and the Cost Disease 12
public policy guiding higher education, particularly with funding allocation decisions.
Reacting to the complaints of parents and students, both Congress and state legislatures
have been seeking ways to control tuition price increases, or at least reduce the rate of
cost and price increases. There seems to be genuine frustration with colleges and
universities, with many believing that higher education institutions are willfully
increasing prices beyond what is necessary. In the latest Public Agenda survey of
attitudes toward higher education, 52% of those surveyed responded that colleges were
more concerned about their "bottom line" than providing a quality educational experience
for students, while 44% responded that waste and mismanagement were major factors in
driving higher education costs (Imrnenvahl& Johnson, 2007). Additionally, 56% of
respondents indicated that colleges and universities could substantially cut their budgets
and lower tuition without sacrificing quality, and 58% either agreed or strongly agreed
that colleges and universities could absorb a much larger number of students without
affecting quality or increasing prices (Immenvahl & Johnson, 2007, p. 25).
These attitudes are not reserved to public opinion surveys, as family concern over
the affordability of a college education has increased, politicians have been expressing
mounting frustration as well. Amendments to the Higher Education Act in the House of
Representatives proposed in November, 2007 reflect a growing bipartisan impatience
with rising tuition prices, with the assumption that colleges and universities are
themselves to blame. Under these proposed amendments, institutions which increase
annual net tuition greater than the percentage increase in the higher education price index
would be required to submit a report to the Secretary of Education describing the factors
which led to its tuition increases, identity the three areas of the institution's current
Higher Education and the Cost Disease 13
budget with the highest increases, submit 3 prior years of filings with the Internal
Revenue Service (or IRS) to the Department of Education, and address planned
institutional actions to reduce its net tuition in the future ("Higher Education Opportunity
Act," 2007).
Meanwhile, skyrocketing costs have been placing unprecedented stress on
institutions and faculty. As cost pressures have mounted, the total number of adjunct
professors and the number of courses taught by adjunct faculty has increased
substantially in the last 30 years, principally as a way to reduce costs (Schuster &
Finkelstein, 2006, p. 40). The economic pressures on faculty have been so extensive that
some have questioned the very hture of the profession as even new permanent faculty
are increasingly hired as non-tenure track positions (Schuster & Finkelstein, 2006, pp.
175-180).
The amendments proposed by Republican representative Castle which called for
those institutions identified as having excessive tuition increases to implement procedures
to cut costs or else face a 10% reduction in federal aid were ultimately withdrawn.
However, Representative Miller, the Democratic Chairman of the Education and Labor
Committee described Castle's amendments as "very tempting," warning, "I hope the
[higher education] community is listening closely to this" adding that the bill's provisions
were "not the end of the story" (Lederman, 2007). State politicians have expressed
similar concern as well.
Federal and state public policy and resource allocation decisions affecting higher
education will hinge on the assessment of the causes for cost and price increases. If these
faster-than-inflation increases are indeed caused by institutional greed or
Higher Education and the Cost Disease 14
mismanagement, or if this is merely believed to be the cause, Congress and state
legislatures may justifiably impose regulatory solutions on institutions, as threatened by
the proposed amendments in the last Higher Education Act, which warned of the loss of
Federal financial aid to institutions which failed to lower price increases.
Moreover, if institutional greed, mismanagement, or skewed priorities are the
cause for tuition price increases, Congress and state legislatures may shift scarce
resources to what are believed to be more pressing public issues. If institutions indeed
have enough funding, but it is merely being misdirected through a warped sense of
priorities, then Congress and states legislatures may feel justified in reducing the relative
resource allocations invested in higher education, believing this may actually help realign
institutional allocations with public policy objectives.
However, if cost and price increases facing higher education are rooted in other
causes, such as those associated with higher education's production function, this may
require very different public policy choices. Such causes, if properly understood, may
actually require a public reinvestment in higher education to strengthen student access
and ensure higher education's continuing function as a primary mechanism of equality of
opportunity.
Definition of Terms
While often linked, the terms cost and price are not synonymous. Cost refers to
the amount institutions spend to provide education to students, and is measured through
expenditures. Price refers to the amount which students are charged and what they pay
for educational services. There are several frequently discussed prices, including sticker
price, price of attendance, and net price. Sticker price refers to the established tuition and
Higher Education and the Cost Disease 15
fees charged by an institution, whle net price refers to tuition and fees excluding
financial aid. When Congress expresses concern about prices, it is usually refemng to
sticker prices (Cunningham et al., 2001). (For a more detailed definition of terms, please
see Appendix A,)
In general, costs are associated with the factors of production; price is the amount
of money charged to students in terms of tuition and fees. However, it should be
recognized that students rarely pay the full cost of the education they receive, even when
paying full tuition, since colleges and universities receive revenue from a variety of
sources to subsidize the costs of education. This general subsidy is defined as the
difference between the average price charged to students and the average cost to the
institution for providing an education to a student, on a per-capita basis.
Higher Education and the Cost Disease 16
CHAPTER 11: REVLEW OF THE RELATED LITERATURE
Researchers attempting to explain rapidly rising cost and price increases in higher
education fall into two broad camps: those seeking causes unique to higher education,
and those attempting to explain cost and price increases as part of a broader phenomenon
potentially affecting all labor intensive industries, known as Cost Disease Theory.
Theoretical perspectives have drawn heavily from economics and include analysis of the
goals associated with higher education institutions as economic firms, the nature of
competition associated with the unique market structure of higher education, the
influence of affluence and market segmentation in relation to cost and price increases,
and the nature of production costs associated with the inputs for higher education. These
theories will be explored in greater detail.
Factors Unique to Higher Education
University Goals, Pursuit of Excellence and the Revenue Theory of Costs: Valuation and Higher Education Expenditures
Goals of Non-Profit Institutions and the Pursuit of Excellence
Most researchers attempting to explain rapidly rising costs and
tuition prices in higher education begin their analysis with the fact that most colleges and
universities are structured as non-profit institutions, which creates significantly different
goals and objectives than those found in profit-oriented firms, a line of enquiry beginning
with Hansmann (1980). Most institutions of higher education are organized as non-profit
enterprises, which stipulate a "non-distribution constraint." While their revenues may
exceed costs, non-profit organizations may not distribute profits to owners or
Higher Education and the Cost Disease 17
stockholders (Hansmann, 1980; Winston, 1996, 1999). While still facing limitations on
possible expenditures, non-profit colleges and universities are not guided by the profit-
maximizing goals of business firms, but are motivated by more complex and less clearly
defined objectives.
Higher education institutions often define their goals in extremely broad terms
such as "the pursuit of excellence" (Winston, 1999), or simply to "be the best" in their
educational, researca and service missions (Clotfelter, 1996). This not only affects how
colleges and universities set their internal priorities, but also the nature of competition
among these institutions. The activities and expenditures associated with colleges and
universities will be affected by the extremely broad goals associated with the majority of
these institutions as non-profit firms. The full implications of the unique structure and
aims of higher education were most fully developed by Bowen (1980), and has since
become known as the Bowen's Law, or the Revenue Theory of Costs.
Bowen (1980) postulated that cost escalation in higher education may not be
strictly related to inherent production costs, but hmged on the potentially unlimited goals
associated with colleges and universities, which he described as "Revenue Theory of
Costs." While colleges and universities are focused on institutional excellence, prestige,
and influence, maintaining these measures of excellence are very expensive. It requires
low faculty-student ratios, higher faculty salaries, high proportions of faculty with
terminal degrees, and release time for research, as well as extensive library holdings,
advanced inkastructure, and equipment. Significantly, the measures on which colleges
and universities assess excellence involve resource inputs, all of which are expensive,
Higher Education and the Cost Disease 18
rather than output and performance measures assessing student learning and personal
development (Bowen, 1980).
Using this as his premise, Bowen postulated that in their quest for prestige and
excellence higher education institutions would spend as much money as they could
acquire in achieving their aims. Non-profit colleges and universities would raise all they
money they could, and spend all the money they raised. The cumulative effect of these
pressures would result in ever increasing expenditures (Bowen, 1980). Massy (2003)
supported Bowen's conclusion, arguing that while institutions justified tuition increases
based on cost increases external to their control, the actual causes for price increases
stemmed from their own choices, based on how they defined excellence.
Zemsky and Massy (1995) also cited the relationship between the decentralized
nature of university decision making and the constant quest for excellence as contributory
factors to the excessive growth of new programs and initiatives, leading to ever
increasing cost commitments. They noted that between 1985 and 1990, college and
university expenditures grew by 3.81% faster than the rate of inflation. Nearly all of this
growth occurred at the periphery of these institutions; for universities, this was primarily
associated with the growth of research centers and institutes, operating near
independently of the university center. However, the expansion of peripheral activities
occurred as well at liberal arts and community colleges, with new degree and continuing
education programs (Zemsky & Massy, 1995).
Clotfelter (1996) and Ehrenberg (1999; 2000) reached similar conclusions as well.
Clotfelter also noted that spending pressures were reinforced by the mechanisms of
shared governance at elite institutions, based on the way they perceived excellence.
Higher Education and the Cost Disease 19
Featuring weak central control, a remarkable degree of freedom accorded to its faculty, and traditions of collegiality in governance, the university lacks any corporate goal other than the pursuit of excellence. When it comes to the research that it undertakes, the university has little to guide it other than an uncompromising devotion to the highest standards inquiry. (Clotfelter, 1996, p. 253)
Massy (2003) recognized there were upward limitations on what colleges can
spend. While providing socially desirable goods, most colleges and universities are
nonetheless heavily dependent on market forces for operating revenue generated from
tuition, which account for 80% percent of the revenue for private colleges, and 28% of
the revenue received by public universities. Thus, universities will attempt to maximize
perceived value subject to two key resource constraints: the demand for their product, and
their available resources (Massy, 2003).
University Goals and Competition for Excellence
A variety of scholars have also noted that the broad and ill-defined goals
associated with achieving academic and institutional excellence have also fostered
extreme competitive pressures among institutions, fueling still greater pressure on
spending to maintain or enhance the relative assessment of institutional quality and
prestige (Clotfelter, 1999; Ehrenberg, 2000; Winston, 1996,2003; Winston &
Zimmerman, 2000). According to this theory, pressure to increase institutional spending
is compounded by the market structure shaping demand for higher education. The market
for higher education is highly decentralized, consisting of thousands of institutions,
which are also segmented into groups that compete on relative selectivity and quality
(Clotfelter, 1999; Winston, 1999,2003; Winston & Zimmerman, 2000). However,
measures of institutional excellence, quality and prestige can only be defined relative to
Higher Education and the Cost Disease 20
other institutions, fueling competition among institutions to be better than others within
their peer grouping (Winston, 1999).
Bowen (1980) recognized that, while driven by desires of excellence, competition
between institutions placed further pressure on institutional spending, providing very
little additional benefits. While Bowen acknowledged that some spending was used to
enhance student quality, in actuality, much of it was used in an attempt to enhance the
relative reputation of a particular institution and attract new donors. This was done
simply to raise the status of one institution relative to another.
Winston (1999) concluded that the pursuit of relative advantage in comparison to
peer institutions had become such a driving force in institutional spending that it had
developed into a "competitive arms race" driving institutional spending beyond mere
necessity:
the notion is that the players have become trapped in a sort of upward spiral, an arms race, seeking relative position; in the case of education, it may, in the extreme, involve expensive 'competitive amenities', that do not produce sufficient benefit to justify their cost directly, but are important to an individual school because others are offering these amenities. (p. 30)
The nature of gaining relative position among a peer group can not only be very
expensive, but also elusive. As Winston (1999) observed, competition may force all
institutions to spend ever greater amounts with little relative effect; "In an arms race,
there is lots of action, a lot of spending, a lot of worry, but, if it's a successful arms race,
nothing much changes. It's the purest case of Alice and the Red Queen where 'it takes all
the running you can do, to keep in the same place"' (p. 40).
Higher Education and the Cost Disease 21
Wealth, Market Structure, and Competition
A new body of research analyzing how institutional wealth shapes the market
niches within which institutions compete has added to the literature attempting to explain
cost increases with industry-specific causes (Winston, 1996, 1999,2003; Winston &
Carbone, 2001; Winston, Carbone, & Lewis, 1998). These researchers have provided a
new perspective on how the competition for student peer quality and the relative wealth
of institutions defines the nature of competition among institutions as well as the
structure of the higher education marketplace. From this perspective, cost and price
increases can be seen as the residual effects of the mechanism of the market structure
associated with the provision of higher education.
Winston (1996) observed that, while still subject to supply-and-demand
constraints, the market for higher education is very different from most others. Most
institutions of higher education receive the revenue to cover their production costs from
two sources, a combination of past and present public and private charitable giving
(collectively characterized as "donative resources"), as well as sales revenues charged to
student "customers". This diversified revenue base consisting of charitable and
commercial sources of revenue allows many colleges and universities to significantly
separate the cost of education from the prices they charge. In fact, Winston (1996)
described the long term separation of the price students pay and the true cost of providing
educational services as one of the most significant features of American higher education.
However, institutions of higher education rely on a highly unique production
technology involving the acquisition of student peer quality in their attempt to provide
Higher Education and the Cost Disease 22
educational quality to their students. This heavily influences the nature of competition
among institutions, with significant impacts on educational cost and price.
Rothschild and White (1995) were the first researchers to recognize the
importance of student peer effects as a core component of the production process of
colleges and universities as commercial entities. Since the business of higher education
was developing human capital, students were a vital input into the production process.
Rothschild and White hypothesized that the presence of particular student customers
impacts the education received by other students. Thus, the educational quality a student
would experience is shaped in part by other students with whom s h e shared hisher
studies (Goethals, Winston, & Zimmerman, 1999).
Unlike nearly every other industry, colleges and universities are only able to
purchase a key production input from the very students who are also the consumers of the
product they are attempting to sell (Winston, 1996; 2003). He defined this unique market
structure as a "customer-input technology." The vital role of student peer quality in the
higher education production process has important implications for competition among
institutions.
Since institutional quality is dependent on the acquisition of student peer quality,
colleges and universities engage in fierce competition in an attempt to attract highly
qualified students. They do this by offering a combination of general and individual
subsidies to students deemed to have desirable academic credentials. The subsidies
offered to a student can be seen as a wage rate for the peer quality based upon hisher
perceived value to the institution (Rothschild & White, 1995; Winston, 1996).
Higher Education and the Cost Disease 23
Winston (1996; 2003) observed there are vast differences in the donative
resources available to institutions, which limits the services, quality of faculty, and the
amount of general and individual student subsidies they can use to attract highly sought
after students to attend their institutions. The great variability in the amount of donative
resources available to institutions across the higher education marketplace enforces a
strict quality and prestige hierarchy; wealthy institutions are able to sell their product at
tuition prices well below the true cost of production, while poorer institutions provide a
significantly smaller subsidy. This affects the number of applicants and student peer
quality an institution is able to attract (Winston, 1996,2003).
Wealthy colleges and universities attempt to insure they will continue to
maximize student peer quality by strictly controlling whom they will allow to consume
their educational services; they can do this by purposely creating excess demand for their
product, hinging on their ability to provide value for students seeking enrollment through
a combination of general and individual student subsidies. Since institutions vary greatly
in their ability to provide subsidies, this reinforces the hierarchy and market segmentation
associated with American higher education. The differential resource base available to
institutions has broad implications for cost and price increases across the higher
education landscape (Winston, 1996,2003).
Institutions that are able to provide significant donative resources are able to
create an excess demand queue for their product, since they are able to provide a
substantial return on a student's tuition investment. They use t h s excess demand queue to
select those students whom they feel will contribute to peer quality. Well-funded
Higher Education and the Cost Disease 24
institutions are able to create a high quality academic program by ensuring they remain
selective in choosing their student body (Winston, 1996).
These factors lead to a heavily segmented marketplace distributed into prestige
bands. For those institutions with access to significant donated wealth, there seems to be
little incentive to control spending in their quest to maximize student peer quality. In fact,
according to the theory, hlgh levels of institutional spending, subsidized by private and
public donated resources, reinforces the value which propels relatively large numbers of
students to apply for the limited number of seats available at selective institutions. This
high-spending relative to price strategy seems to create and reinforce the selectivity
associated with the most prestigious institutions.
Merit Aid, Tuition Discounting, and Rising Costs
A significant body of research has identified increases in institutional non-need
based aid as a rising component of institutional budgets, especially at private institutions
(Clotfelter, 1996; Hauptman, 1997; Mulugetta, Saleh, & Mulugetta, 1997; Wellman,
2001). For most institutions, these expenditures are not covered by institutional
endowments, but are associated with the growing phenomenon known as tuition
discounting.
Tuition discounting is the practice of purposely charging lugher tuition sticker
prices, while providing significant discounts to students with highly desirable
characteristics, such as those with high academic credentials. Institutions engaging in
tuition discounting charge those students with less desirable academic credentials a
higher tuition price than highly sought after students, while those students who barely
meet academic criteria for admittance are charged the full tuition sticker price. Thus,
Higher Education and the Cost Disease 25
lower ranked students help subsidize the education of their more academically gifted
peers (Baum & Lapovsky, 2006). This practice is often utilized by less highly selective
institutions which do not have large endowments to provide scholarships or offer
substantial general subsidies to all students. Instead, the tuition discount is targeted to
specific students to entice them to enroll at the institution.
The growing reliance on tuition discounting is well documented. As described in
a 2001 Carnegie Commission study (Wellman, 2001), the use of tuition discounting has
been accelerating, and has been classified as one of the most important causes for
increases in tuition sticker prices at private colleges and universities. The spiraling
increase in tuition discounting in an effort to increase indices of student quality has even
been categorized as an "arms race," increasingly permeating less selective segments of
the higher education marketplace (Wellman, 2001).
Some researchers have identified institutional merit based aid as one of the fastest
growing components of institutional expenditures. In their Study of College Costs and
Prices, 1988-89 to 1997-98 conducted for the NCES, Cunningham et al. (2001) found
that from 1988-89 through 1997-98 expenditures for institutional aid at public institutions
grew at an average annual inflation-adjusted rate of 8.1% at research/doctoral institutions,
7.7% at comprehensive institutions, and 6.8% at 2-year institutions. Institutional aid at
private institutions grew at even faster rates: between 1988-89 and 1995-96, expenditures
for institutional aid at private institutions grew at an average annual inflation-adjusted
rate of 8.5% at bachelor's institutions, 8.7% at researchldoctoral institutions, and 10.2%
at comprehensive institutions (Cunningham et al., 2001).
Higher Education and the Cost Disease 26
Even these statistics may mask the true extent of growth in merit aid (compared to
increases in all institutional aid) among institutions of higher education. McPherson and
Shapiro (2002) suggested that prior research has actually underestimated the full extent
that colleges have shifted their institutional aid from need to merit based aid; they
examined the merit and need-based awards offered to sample of 7,000 full-time
dependent students derived from the National Postsecondary A d Survey in 1995-96, who
had taken the SAT exam.
McPherson and Shapiro (2002) discovered that relatively few students receive an
award explicitly defined as merit or non-need based. Of their sample, only 4% of students
at public and 15% of students at private colleges received an award classified as merit or
non-need, while 22% of the students from public colleges and 52% of students from
private colleges received awards designated as need-based (McPherson & Schapiro,
2002). However, they found that even within aid classified as need-based, SAT scores
had a significant impact on the size of the award a student received. Low income students
entering a public college with a high SAT score received an average of $1,255 in need-
based aid, compared to an average grant of $904 for low income students with mid-range
SAT scores, and just $565 for low income students with SAT scores categorized as low.
The differences were even greater at private colleges; low income students with
high SAT scores received an average institutional need-based grant of $4,741, while low
income students with low SAT scores received an average award of $1,028 (McPherson
& Schapiro, 2002). The high correlation of not only merit based grants but also aid
classified as need-based to SAT scores seems to indicate that colleges and universities are
Higher Education and the Cost Disease 27
massively shifting their institutional aid resources toward merit aid as a way to recruit
students perceived to be high-achieving (McPherson & Schapiro, 2002).
It seems clear that expenditures on institutional grant aid, particularly merit-based
aid often based on tuition discounting, has been one of the fastest growing components of
institutional budgets. While the increasing role of merit-based aid in institutional
expenditures has been analyzed as a distinct trend in the college cost literature, the work
of Winston (1996,2003) provides an opportunity to place the role of merit-based aid
within the broader context of higher education's market structure and its unique
customer-input technology.
The theory of the role of the customer as a production input implies that students
provide the critical element of peer-quality required by institutions in their attempt to
provide a rich educational experience (Rothschild & White, 1995; Winston, 1996).
However, the ability to entice highly sought after students to enroll at a particular
institution is tied to the institution's ability to offer value to students by providing a large
subsidy relative to the cost of attendance (Winston, 1996; 2003). The most elite
institutions with the largest donative resources accomplish this by providing a large
general subsidy to each student who enrolls. Institutions unable to provide a large general
subsidy can attempt to selectively offer individual students generous merit aid awards as
an inducement to enroll. Winston's economic model describing the role of students in
providing peer-quality and the competition it engenders can help provide a unifying
theme to the rise of merit-based aid and tuition discounting as a direct consequence of
higher education's market structure, and the competition for student peer-quality.
Higher Education and the Cost Disease 28
Inherent Production Cost Factors
As described in the previous section, a significant body of research has focused on
the relationship between the aspirations of non-profit colleges and universities and the
competition it engenders to explain rapidly rising costs and tuition sticker prices.
However, many researchers remain convinced that this is due to the specialized
production factors associated with higher education or, at least, exacerbated by these
inherent production costs. Put another way, competition may be fueling institutional
expenditures, but the factors of production on which institutions compete may be
disproportionately expensive, leading to faster than average cost increases, which may
then be passed on to students in the form of higher tuition prices. The empirical,
descriptive studies mandated by Congress have focused on this issue, attempting to
identify whether there are unique "cost dnvers" in higher education (Cunningham et al.,
2001; Jones, 2001). Other researchers have delved into this potentially powerful
explanatory factor as well.
In their Study of College Costs and Prices, 1988-89 to 1997-98 conducted for the
National Center for Educational Statistics (NCES), Cunningham, Wellman, Clinedinst,
Meriosotis, and Carroll (2001), attempted to assess the relationship between cost
increases in the production function of higher education and their relationship to price
increases. This study identified a number of higher education cost drivers, which include
facilities, institutional aid, technology, regulations, mission and discipline, faculty
compensation and workload policies, as well as class size. Significantly, other than the
cost of regulatory mandates, these included nearly all the core inputs used to enhance
institutional excellence (Brennan, 2001; Cunningham et al., 2001). Most of these cost
Higher Education and the Cost Disease 29
drivers have been identified by other scholars including Bowen (1980), Clotfelter (1996,
1999) and Ehrenberg (1999,2000) as production factors which consistently outpace
increases in the Consumer Price Index, leading to rapidly rising costs.
While driven by desires of institutional excellence, Bowen (1980) acknowledged
that costly resource inputs were fundamental to the Revenue Theory of Costs;
significantly, excellence was not measured by outcomes assessments of effectiveness in
educating students, but in terms of resource inputs which were expensive to maintain and
enhance.
In his case study of Cornell, Ehrenberg (2000) supported Bowen's conclusions.
m l e pressure for spending was based on the pursuit of excellence, Ehrenberg described
the nature of these costs as the very expensive resource inputs which the pursuit of
excellence requires. Elite private institutions seek to deliver the best quality
undergraduate and graduate education they can; they will seek as many faculty members
as they can to enable small class sizes and allow greater research in cutting edge fields of
study. Extending knowledge by engaging in research is also very expensive, requiring not
only faculty, but also state of the art facilities, laboratories and equipment. While a great
deal of this research is supported by government, corporation and foundation grants or
contracts, much of the research is in fact sponsored by universities themselves in the form
of reduced teaching loads afforded to faculty engaged in research. There is a constant
pressure to reduce faculty teaching loads even further to provide opportunities for even
more research. This creates what Massy (1996,2003) described as a "ratchet effect,"
leading to the gradual decline in teaching loads and even greater educational costs.
Higher Education and the Cost Disease 30
Cost Disease of the Service Sector
The education of students would be impossible without faculty able to teach. The
specialized labor of highly educated faculty is one of the most important inputs in the
higher education production process. However, some researchers have identified labor
costs as one of the factors of production whch has increased faster than the CPI; a special
literature has developed around the problem of cost increases in heavily labor intensive
industries which warrants closer study (Baumol, 1967, 1993; Baumol, Blackman Batey,
& Wolff, 1985; Baumol & Bowen, 1965,1966).
Baumol and Bowen (1965, 1966) were the first theorists to propose that heavily
labor intensive industries will face faster than average cost increases compared to those
associated with the manufacturing sector. This literature has been extended to potentially
explain faster-than-inflation cost increases and their relationship to tuition and fee
charges in higher education, and has since become known as the "Cost Disease of the
Service Sector."
Baumol and Bowen's (1965, 1966) original research focused on cost increases in
the performing arts; they began by noting the puzzling paradox that, while the median
income of the majority of performing artists was much lower than most other
occupations, the cost of ticket prices seemed to be rising at a pace much more rapid than
general inflation. They came to the remarkable conclusion that the explanation centered
on the relatively static rate of productivity increases in the performing arts compared with
the manufacturing sector, leading to faster-than-inflation increases, not just in the
performing arts, but in other service sector related jobs as well (Baumol & Bowen, 1966).
Higher Education and the Cost Disease 31
In their initial iteration of their theory, which Baumol and his colleagues later
refined, Baumol and Bowen (1965; 1966) hypothesized that the economy could be
divided into two broad sectors, a progressive sector which was able to successfully
leverage technological innovation, capital, and economies of scale to increase output, and
a "stagnant" sector, which was only able to increase productivity inconsistently and
sporadically if at all. The classification of the economy into these sectors was not
arbitrary, but based on the technological structure of a particular industry.
The key distinction was the role of labor in the production of goods and services
within particular industries; in certain industries, particularly manufacturing, labor was
only an input into the production process whlch was incidental to the utility of the
finished product. For example, when purchasing a computer, consumers are not
necessarily concerned with the amount of labor utilized in the production process, but
only about the price and quality associated with the finished product. As long as quality is
unaffected or actually improves, and prices remain stable or actually decrease,
manufacturers can leverage new technology to reduce the amount of manpower required
in the production process (Baumol, 1967).
Baumol and Bowen contrasted these manufactured goods with service sector jobs
where labor itself is the end product. In these activities, the quality of the service
provided is directly related to the amount of labor allocated to the task. Their now classic
example is the production of a live half-hour music quintet; this requires the expenditure
of 2 % person hours to produce, and any attempt to increase "productivity" by deceasing
the amount of labor expended simply leads to a reduction of quality. Significantly,
Higher Education and the Cost Disease 32
another example they cited involved teaching, in which class size is used to measure the
quality of education provided to students (Baurnol, 1967).
The theory proposed by Baumol and Bowen's model has important implications
for the relationship between wages and costs in the productive and stagnant components
of the economy. They hypothesized that wages in the two components of the economy
fluctuated in tandem, while it was possible for wages in one sector to lag behind, fluidity
in labor markets meant that this could not occur forever, and that the wage rate of those in
both the productive and stagnant sectors increased at approximately the same rate
(Baumol, 1967). In general, wages in the manufacturing sector would rise as rapidly as
productivity increases per person hour; however, since labor markets are fluid, wage
increases in the static sector would generally mirror increases associated with the
productive manufacturing sector as well.
The implications for costs in the two sectors were profound. Since wages in both
the productive and stagnant sectors of the economy increased at approximately the same
rate, which could be mitigated by productivity increases only in the productive sectors of
the economy, costs in the nonproductive labor-intensive sector would rise faster than
those in the productive sector, which Baumol(1967) hypothesized later would rise
"cumulatively and without limit".
The Theory of a Cost Disease has important implications for higher education.
The theory postulates that the issue of cost increases in higher education may be part of a
broader phenomenon related to all service sector jobs, including the performing arts,
health care, postal services, and even automotive repair, among others (Baumol, 1967,
1993; Baurnol & Bowen, 1966). The higher rate of cost increases in the service sector
Higher Education and the Cost Disease 33
centers on the differential rate of productivity increases between service sector, and
manufacturing sector jobs. Higher education is part of a class of activity where the
production process lacks effective methods of standardization and automation (Brennan,
2001). Moreover, for many services, particularly education, it is extremely difficult to
reduce the amount of labor involved in providing services, since quality of service may
be directly related to the amount of labor provided (Baumol, 1993). Therefore, the quality
of service usually begins to deteriorate if less time is allocated by doctors, auto
mechanics, and college and university faculty in performing their jobs (Baumol, 1993)
In contrast, labor-saving technology can be implemented in the manufacturing
sector without a reduction in quality. Consequently, it has been much easier to increase
productivity in manufacturing as opposed to jobs in service-related industries (Baumol et
al., 1985).
The difficulty of introducing labor saving technology will not allow wages of
service sector personnel to be offset by larger productivity gains to the same extent as in
manufacturing industries. Thus, service sector jobs will grow more expensive, relative to
manufactured goods, even after inflation is accounted for (Baumol, 1967; Baumol &
Blinder, 1985).
Even as Baumol and Bowen focused on the performing arts, they were quick to
point out the implications of their theory for higher education as well:
It is evident that the foregoing analysis is applicable to many services other than the performing arts.. .it helps to explain the financial problems of higher education, which have received even more publicity than those of the performing arts. Education, like the arts, affords little opportunity for systematic and cumulative increases in productivity. The most direct way to increase output per hour of teaching - an increase in the size of classes - usually results in a deterioration of the product which is unacceptable to much of the community. Thus, the financial problems
Higher Education and the Cost Disease 34
which beset education are, at least in part, another manifestation of the fundamental relationship between productivity and costs which is so critical for the living arts. (Baumol & Bowen, 1966, p. 171)
Baumol later expanded on the broader implications of this theory, including
higher education. In a 1967 article entitled, "Macroeconomics of Unbalanced Growth:
The Anatomy of Urban Crisis", Bawnol observed that the challenges to urban growth and
renewal centered at least in part on the fact that the costs of government services were
rising faster than general costs due to the relative difficulty of increasing productivity in
services associated with municipal administration.
In the same article, Baumol again postulated that the cost disease would continue
to plague higher education as well:
Higher education is another activity the demand for whose product seems to be relatively income elastic and price inelastic. Higher tuition charges undoubtedly impose serious hardships on lower-income students. But, because a college degree seems increasingly to be a necessary condition for employment in a variety of attractive occupations, most families have apparently been prepared to pay ever larger fess to institutions in recent years. As a result, higher education has been absorbing a constantly increasing proportion of per-capita income. And the relatively constant productivity of college teaching leads our model to predict that rising educational costs are no temporary phenomenon-that they are not a resultant wartime inflation which will vanish once faculty salaries are restored to their prewar levels. Rather, it suggests that, as productivity in the remainder of the economy continues to increase, costs of running the educational organizations will mount correspondingly, so that whatever the magnitude of the funds they need today, we can be reasonably certain that they will require more tomorrow, and even more on the day after that. (Baumol, 1967)
The theory of a cost disease has generated considerable controversy, and many
researchers have expressed skepticism since it was proposed. Much of the early
criticism within the higher education community accepted that overall costs and
tuition price increases were indeed rising faster than the consumer price index, but
Higher Education and the Cost Disease 35
questioned whether this was driven by the core functions associated with the
education of students. Brovender (1974) and James (1978, Spring) suggested that
actual per-unit credit hour costs have not increased as rapidly as the cost disease
suggests and there may indeed be true economies of scale in delivering educational
services if one factors in the change in output mix associated with the allocation of
time college and university faculty spend on the triad of functions, education,
service, and research.
More recently, Massy (1996) has argued against the presence of a systematic
cost disease, suggesting that it does not explain the preponderance of cost increases
in higher education. While acknowledging that between 70% to 80% percent of
college and university operating expenses are related to personnel and benefit costs,
he concluded that the labor-intensive nature of higher education can only account for
one percentage point above the normal growth in the CPI. He concluded that while
faculty productivity gains may indeed be limited by the nature of student-faculty
ratios, actual faculty personnel costs account for only 50% to 70% of educational and
general expenses of a college or university budget. Since productivity gains in the
United States tend to average at an annual rate of between 1 %% and 2%, the rest of
university activities should be subject to similar productivity gains (Massy, 1996).
Although focused on analyzing cost increases in government services, and not
those associated with higher education, Fenis and West (1996) concluded that, while
slower than the manufacturing sector, productivity increases in the labor sector were
still possible. They estimated that slower productivity growth in government services
accounted for two-thirds of the increased cost in government services, while the
Higher Education and the Cost Disease 36
remaining third was due to increases in real wages, based on productivity increases.
Thus, some effects of a cost disease could be somewhat mitigated through real
productivity increases, although these were lower than those associated with the
manufacturing sector.
The theory of the cost disease, if true, has profound implications for higher
education. As indicated above, in education, since the faculty-student ratio remains
relatively constant, and in fact, is used as a measure of quality, the productivity
growth per contact hour is often effectively zero, thus, an increase in nominal
salaries for faculty will lead to an increase in the per-capita cost of providing
education to students.
As described, the cost disease will affect those industries whose expenditures are
disproportionately comprised of personnel costs. While it is difficult to measure the
percentage of labor-related costs in higher education, it is significant: The best estimates
indicate that higher education costs are disproportionately salary-driven, 5 accounting for
70% to 80% of an institution's operating budget (Massy, 1996, p. 53).
Other researchers have confirmed the importance of personnel costs in the cost
structure of higher education (Brennan, 2001, p. 15; Clotfelter, 1996, 1999; Ehrenberg,
1999,2000), often describing a compounding effect between personnel costs and the
nature and structure of university goals and the pursuit of excellence as colleges and
universities attempt to hire quality faculty to maintain or enhance institutional excellence,
or reduce teaching loads to allow faculty to engage in important research activities.
In h s assessment of cost increases in higher education, Johnstone (2001) also
acknowledged that colleges and universities suffered from a general "productivity
Higher Education and the Cost Disease 37
immunity," based on their inability to successfully substitute capital for labor. Moreover,
the problem of rapidly rising college costs appears to be a world-wide phenomenon,
leading many governments, even those in Europe where the concept of free or low cost
public education is a deeply embedded public value, to attempt to find new pattems of
revenue sharing to ease pressure on public expenditures (Johnstone, 2003).
The Theory of the Cost Disease is distinct among the explanations attempting to
account for cost and price issues in higher education. All other causal explanations rely
on industry-specific factors unique to higher education. Cost Disease Theory offers an
explanation which places cost increases in higher education as part of a broader
phenomenon affecting all labor-intensive industries. The broader focus attempting to
place higher education costs within the context of other service sector industries also
differentiates the type of evidence used to assess the validity of Cost Disease Theory in
explaining higher education costs. Unlike theories explaining cost and price increases
which are higher education industry-specific, evidence relating to the existence of a cost
disease requires that higher education costs must be assessed against other labor-intensive
industries to determine whether and to what extent labor costs shape cost pattems in these
industries. The comparison of higher education costs to costs in other labor intensive
industries has yielded some very important findings, and has refocused attention on Cost
Disease Theory in explaining cost and price increases in higher education (Archibald &
Feldman, 2008).
While originally at the margins of the higher education cost debate, the Theory of
the Cost Disease has gained tremendous currency, and is now considered one of the two
Higher Education and the Cost Disease 38
most common causes used to explain faster than-inflation increases in higher education
costs, along with Bowen's Revenue Theory of Costs (Archibald & Feldman, 2008).
Among single theory explanations developed to explain cost and price behavior of
colleges and universities, Cost Disease Theory seems to provide the most explanatory
power. While all the factors analyzed in this study are most likely associated with faster-
than-inflation costs and associated price increases higher education, including the goals
of non-profit institutions and the competition among institutions, there is considerable
evidence that the majority of hlgher education cost increases are directly associated with
the labor-intensive nature of the industry and the inability to easily substitute technology
in the education of students without sacrificing educational quality.
After the initial studies developed by Baumol and Bowen (1965, 1966), a great
deal of empirical analysis was conducted which provided additional support that labor-
intensive industries are subject to a cost disease (Archibald & Feldman, 2008; Baumol,
1967, 1993; Baumol et al., 1985; Brennan, 2001; Clotfelter, 1996; Fems & West, 1996;
Johnstone, 2001,2003). This impressive body of descriptive analysis has supplemented
economic theory associated with the cost disease to provide strong evidence that
disproportionately labor intensive enterprises, both profit-oriented and non-profit in their
ownership structure, face cost increases that are higher than the general economy. In fact,
as Clotfelter (1996) indicates, the rate of cost increases in higher education seems to
mirror those associated with other professional services, even those in for-profit firms.
Between 1980 and 1992, the earnings of full professors increased an at an average annual
rate of 1.3% above inflation (measured by the CPI), while the salaries of non-supervisory
Higher Education and the Cost Disease 39
attorneys increased at a annual rate of 1.1 % above the CPI, chief legal officers at 1.4 %,
and physicians at a rate of 2.3% above the CPI (Clotfelter, 1996).
This consistent pattern can be seen in other industries, not only in the United
States, but in other industrialized countries as well. Analyzing OECD data from 1960 to
1987, Baumol(1993) discovered that the growth rate of health care costs have
consistently outpaced the GDP deflator index in nearly all OECD counties, independent
of whether they were part of a private or public health care delivery system. Similarly,
the relative price of automotive repairs and auto insurance has increased at a rate higher
than the CPI, and at rates similar to those for education and health care (Baumol, 1993).
It also seems that potential issues associated with a cost disease are not limited to
higher education. Examining the costs of primary and secondary education in a selected
group of OECD countries, Gundlach, Wossman and Gmelin (2001) observed that low
productivity growth in schooling was responsible for cost increases which exceeded
inflation.
The large number of descriptive studies analyzing a variety of heavily labor-
intensive industries provides the strongest support for the notion that a disproportionate
percentage of higher education cost increases are directly related to its labor-intensive
nature. This explanation contrasts most starkly with Bowen's Revenue Theory of Costs
(1980). Bowen's theory describes the majority of higher education cost as deriving from
the goals and aspirations of colleges and universities as non-profit entities, attempting to
maximize quality as opposed to the more traditional goal of profit-oriented entities of
maximizing profit. This would imply that large cost increases such as those observed in
higher education should also be limited to other non-profit entities. However, this does
Higher Education and the Cost Disease 40
not appear to be the case. The higher rate of cost increases found in higher education
seem to be found in labor-intensive profit-oriented firms as well, providing strong,
empirical evidence that the single leading cause of faster-than-inflation cost increases
associated with higher education is the labor-intensive nature of the industry, and not the
goal structure of non-profit colleges and universities.
Literature Review - Prior Methodology
While a variety of scholars have attempted to determine whether labor-intensive
industries are affected by a cost disease, very little research has been conducted
specifically relating to higher education. Researchers attempting to assess the existence of
a cost disease have approached the problem in innovative ways, while constrained by the
data available for analysis. A series of innovative studies have been conducted comparing
both worker productivity and costs in service sector jobs compared to those in the
manufacturing sector. All of these prior studies have utilized time series data and trend
analysis to discern the extent to which prices in labor-intensive industries follow a
trajectory different from those associated with price increases in the general economy or
those associated with the manufacturing sector. While not focused on education in
general, these studies nonetheless have direct bearing on the issue of cost and price
increases in higher education, in addition to wide-ranging implications for the general
economy.
Baumol and Bowen (1965, 1966) began this analytical kamework in a series of
works analyzing price increases over time associated with the live performing arts,
compared to salaries of performing artists. Times series analysis examining the
Higher Education and the Cost Disease 41
relationship between costs and productivity increases was later extended to examine
trends in local government costs in relation to worker productivity in government service
sector jobs, and the impact of stagnant productivity in government senices on municipal
spending pressures and the emerging urban crisis of the 1960's (Baumol, 1967).
Trend analysis has been used to analyze issues of cost increases in a variety of
service sector industries and the impact of productivity growth on a number of important
public policy issues. Baumol(1985) utilized the Bureau of Economic Analysis' National
Income and Product Account data to assess average annual rates of productivity growth
for various sectors of the economy between 1947 and 1976. Using cost and deflator index
data from 1965 to 1985, Baumol(1993) analyzed increases in real educational (K-12)
costs for the United States and other selected countries associated with the Organization
for Economic Cooperation and Development (OECD), as well as using this method to
analyze health care costs for OECD countries from 1960 to 1987. The analysis of costs in
relation to productivity increases across labor sectors has provided a vital insight into the
puzzle of cost and price increases in higher education, which will discussed in more
detail below.
Since the effects of a cost disease are based on the limitations of leveraging
technology to increase productivity in labor intensive industries, researchers assessing the
effects of a cost disease have attempted to find evidence by assessing price changes
across time among various industries. Increases in price should not be random; instead,
those industries which are heavily labor intensive should witness substantially greater
cost increases than those associated with the manufacturing sector.
Higher Education and the Cost Disease 42
Archibald and Feldman (2008) successfully utilized time series data maintained by the
Bureau of Economic Analysis (BEA) National Income and Product Accounts (NIPA) to
compare cost increases in higher education with other industries. As one of NPA's
datasets, the BEA has maintained price indices for 103 industries, the majority of which
has been continuously collected since 1929. They used trend analysis associated with the
NIPA dataset to analyze real price changes in higher education with other industries from
1949-50 to 1995-96, concluding that higher education cost rose at approximately the
same relative rate as other industries in the personal services. They were, however,
hampered in reaching more definitive conclusions. While N P A data aggregates the
product categories into sectors such as services, Archibald and Feldman noted that the
categories utilized by NIPA did not necessarily correspond with sectors of the economy
defined by cost disease theory. The Cost disease theory defined by Baumol and Bowen
predicts that prices associated with labor intensive personal services should experience
price increases which are higher than higher than those associated with the manufacturing
sector. However, NIPA classifies goods and services in very gross categories, which do
not account for the role of labor in the underlying production technology. Since they
could not find a way to define product categories into exact classifications which would
correspond to the role of labor in production, they determined they could not provide a
statistical test to determine the significance of their analysis, which they acknowledged
also limited the conclusions they could draw. However, Archibald and Feldman's
approach provides another framework for assessing whether higher education costs are
unique, or whether similarities exist between service sector industries.
Higher Education and the Cost Disease 43
Many of the limitations in their initial dataset noted by Archibald and Cox
have been corrected in the latest 5-year revision of the NIPA classification system,
completed in September, 2009. As part of a 5 year review and revision by the Bureau
of Economic Analysis (BEA), the definitions associated with NIF'A product
categories were modified to more closely match definitions associated with durable
goods, non-durable goods, and services (McCully & Teensma, 2008). This allows
much more meaningful comparisons across aggregate categories defined by the
BEA.
Moreover, with so few researchers conducting primary research on the effects of a
cost disease specifically within higher education, it seems that an important dataset has
been ignored in analyzing a relationship between a cost disease and college and
university cost and tuition prices; this is the higher education price index (HEPI). HEPI
was first developed in 1961, and has been collected ever since, although the method used
to calculate the index was radically changed in 2002. From 1961 until 2001, the
aggregated HEPI index was based on core price data for over 100 subcomponents,
representing a typical market basket of goods and services consumed by colleges and
universities. Beginning in 2002, the computation has been based on a regression formula,
based on eight core components, which represent 79.6 percent of prior weighted HEPI
index (Commonfund Institute, 2004). Deconstructing the components of the HEPI index
may assist in analyzing the extent to which labor costs have influenced higher education
costs, which will be discussed in more detail below.
Higher Education and the Cost Disease 44
CHAPTER 111: METHODS
Conceptual Framework
This study utilizes the fixmework developed by Archibald and Feldman (2008),
with some modifications. The study conducted by Archibald and Feldman utilized
comparative price indices associated with the National Income and Product Accounts'
Personal Consumption Expenditure by Product Category to compare price increases in
higher education relative to other product categories. Using time series analysis,
Archibald and Feldman sought to test the extent to which industries represented in the
National Income and Product Accounts followed a price path similar to those associated
with higher education. Rather than analyzing price rises across the entire series, they
divided their dataset, spanning the years from 1949-50 through 1995-96 into 11 distinct
segments of 4 years each. For each of the associated 4-year time segments, they
calculated a measure of real price increases (Archibald & Feldman, 2008, p. 283).
They then calculated an "absolute difference" between the price index associated
with each of the NIPA product categories and the price index for higher education for the
same period. A price index which followed the same price trajectory as higher education
would thus have an absolute difference of zero, while the absolute difference would
become relatively greater as the differences between a particular product category and the
higher education real price differed during the period under study (Archibald & Feldman,
2008).
Finally, Archibald and Feldman calculated an index reflecting the real price
change for each product category from 1949-50 through 1995-96, along with a mean
Higher Education and the Cost Disease 45
absolute deviation, the mean of their calculated absolute dfference, spanning the entire
period from 1949-50 through 1995-96 (Archibald & Feldman, 2008).
Archibald and Feldman sorted the list of product categories by the mean absolute
deviation. They noted that 18 of the top 20 product categories with the highest mean
absolute deviation were services, out of a total of number of 69. Of these categories, 13
were durable goods, 17 were classified as non-durable, and 39 were services.
However, as described above, at the time they conducted their test, the NLPA
classification system did not always properly categorize product categories according to
the intensity of labor as a production input. They conducted a simple probability test,
based on the chance that so many service product categories would randomly be
associated with the highest mean absolute deviation scores. They concluded that the
chance of randomly drawing 18 services among the top 20 categories experiencing the
greatest price increases among a distribution containing 39 services and 30 types of other
goods was ,003. They concluded this was sufficient to reject their null hypothesis that
those product categories containing price increases most resembling higher education
would be randomly distributed (Archibald & Feldman, 2008, p. 283).
For this study, the researcher utilizes price indices in two complementary ways, in
an attempt to both deconstruct cost pressures within higher education, as well as assess
the extent to which higher education cost and price increases mirror those associated with
other labor-intensive enterprises.
Ideally, evidence that a cost disease affects higher education should involve two
distinct components; first, there should be evidence that, within higher education,
personnel costs have been a significant driver of total costs. This can be done by
Higher Education and the Cost Disease 46
deconstructing higher education's industry specific price index to determine the factors
most responsible college and university cost increases. The Higher Education Price
Index, first established in 1961, provides valuable data to assess these industry-specific
cost dnvers. An analysis of this data will help determine the extent to which labor costs
fuel total higher education costs over time.
However, by itself, this is insufficient. Industry-specific explanations of hgher
education cost can also potentially explain these increases. As described above, the
Theory of a Cost Disease is the only theory attempting to explain tuition price increases
based on factors which are not solely unique to institutions of higher education, but based
on forces affecting all labor-intensive industries. This also implies that evidence for a
Cost Disease affecting higher education should also be evident in other industries of the
"stagnant" sector hypothesized by Baumol and Bowen.
Fully testing the extent to which a cost disease plagues hgher education also
requires a comparison of price increases in higher education relative to other industries
over time to determine whether the price increases experienced in higher education
mirror those associated with other labor intensive industries and, indeed, whether and to
what extent there is any evidence for differences in the rate of price increases between
manufacturing and labor-intensive industries.
Thus, in addition to assessing cost drivers within higher education, the researcher
also analyzes higher education cost and price increases in a comparative context against
other labor-intensive industries as well as manufacturing sectors of the economy.
Industry-specific explanations of higher education cost and price increases focus on the
distinct incentive structure facing the majority of colleges and universities as non-profit
Higher Education and the Cost Disease 47
entities. If these explanations are correct, researchers should not be able to uncover any
particular pattern in price increases between different industries. However, if cost disease
theory is valid, there should be discernable patterns of price behavior between labor-
intensive and predominantly manufacturing based industries, whether or not the firms
associated with a particular industry are profit-oriented or non-profit based entities.
Moreover, researchers would also expect to find higher education price increases more
closely match increases associated with other labor-intensive enterprises.
To determine the extent of a relationship between a cost disease and higher
education cost increases, components of the Higher Education Price Index are used to
deconstruct the cost pressures within higher education's production function to determine
the extent to which labor costs are responsible for driving total industry costs. Second, the
researcher utilizes national price indexes available from the National Income and Product
Accounts' "Personal Consumption Expenditure by Product Category" for all available
industries to compare cost and price increases in higher education to those in
manufacturing and other labor-intensive industries. This provides a comparative context
to determine the extent to which cost and price increases in higher education increases
mirror price increases associated with other labor-intensive industries, as well as the
extent to which price increases differ between industries in the manufacturing as opposed
to those in the service sector. These price indices will be described later in greater detail.
A limitation acknowledged by Archibald and Feldrnan was their inability, due to
limitations in the data, to completely isolate product categories associated with the
personal services, and, thus, to provide more a robust statistical test of how cost increases
in higher education corresponded to other service sector industries. While the NIPA
Higher Education and the Cost Disease 48
dataset available to them during their study did provide an aggregation for a group of
activities it classifies as "services," this grouping was based on an older classification
system which included some categories of services which are not truly personal services.
Revisions in the classification of the Personal Consumption Expenditures dataset,
completed in September 2009, have created a much more useful classification system to
distinguish between durable goods, non-durable goods, and service industries, also
enhancing the researcher's ability to assess the extent to which differences in price
increases exist between labor-intensive and manufacturing industries.
Research Methods
This study relies on historical time series price indices to analyze the extent to
which a cost disease influences cost and price increases within higher education. Two
different cost and price indices, the Higher Education Price Index, created in 1961, and
the National Income and Product Accounts' Personal Consumption Expenditures by
Product Category will be used to determine (a) the extent to which personnel costs within
higher education are responsible for driving higher education costs, and (b) the extent to
which higher education cost and price increases are similar to those associated with other
labor-intensive industries.
The two price indices which are used cover different timelines and are indexed to
different base years. In order to provide meaningful comparative data, a decision must be
made as to the proper timeframe under examination and a method to create a common,
standardized index to allow analysis between indices. The higher education price index
was created in 1961, and currently uses 1983 as a base year index. The Index for Personal
Consumption Expenditures by Type of Product currently contains 72 detail product
Higher Education and the Cost Disease 49
categories, using 2005 as a base year. The researcher analyzes both HEPI and PCE data
using a base year of 1961, the date for the creation of the Higher Education Price Index,
which allows a comparative context both to analyze the extent of a relationship between
labor costs and overall cost increases within higher education, as well as comparing cost
and price increases in higher education to other product categories. This requires re-
indexing both the HEPI and PCE data to a 1961 benchmark year, which also provides 69
detail PCE product categories. The two data sources, their uses and limitations will be
discussed in further detail.
Data Sources
As described in the previous section, the researcher utilizes two distinct datasets:
(a) price indices associated with the National Income and Product Accounts' Personal
Consumption Expenditure by Product Category and (b) the Higher Education Price
Index.
Role and Uses of Price Indexes
A price index is simply an attempt to measure changes in prices for a commodity
or a market basket of goods over time. This will take into account the potential erosion of
purchasing power due to inflation. For example, the most widely used price index, the
Consumer Price Index, attempts to measure the average cost for a standard market basket
of goods over time. For the CPI, this includes typical goods and services used by
families, and includes the price of food, clothing, shelter, fuel, transportation, and the cost
of medical care (Commonfund Institute, 2004, p. 3). Since the same market basket is
Higher Education and the Cost Disease 50
assessed across time, this creates an average measure of price increases over time, and
thus, a measure of the inflation rate.
It is important to remember that the inflation measured by a price index differs
from total increases in expenditures for a particular market basket. Expenditures are the
product of the amount of items purchased times the current price. The total net price
spent on a product will reflect both increases in consumption, along with any potential
price increase due to inflation.
In addition to the consumer price index, a variety of different price indices have
been created to determine the extent to which service and commodity prices have
increased over time, including the Personal Consumption Expenditure by Product
Category and the Higher Education Price Index.
Personal Consumption Expenditure by Product Categoly
The most comprehensive series of price indices have been maintained by the
Bureau of Economic Analysis, as part of the National Income and Product Accounts
(NIPA). NlPA and the Personal Consumption Expenditure indices were originally
developed during the 1930's, to provide metrics to assist New Deal officials in
formulating policies to cope with the Great Depression. The accounts and metrics were
extended during World War I1 to assist government planners manage economic resources
for war production (Marcuss & Kane, February, 2007).
N P A measures changes in the Gross Domestic Product (GDP), as measured
through Personal Consumption Expenditures (PCE). PCE indices measure goods and
services consumed by final purchases within the US economy. This includes households,
non-profit institutions serving households residing in the US, as well as purchases by the
Higher Education and the Cost Disease 5 1
US government. The majority of PCE purchases are comprised of new goods and
services provided by private business consumed by households. In addition to quantifying
how much of household income is spent on the consumption of goods and services, PCE
also provides a mechanism measuring the types of goods and services purchased (Bureau
of Economic Analysis, 2009).
During 2009, the Bureau of Economic Analysis conducted a comprehensive
revision of the Personal Consumption Expenditure indices, creating a new classification
system to more fully reflect changes in consumption patterns created by changmg
demographics, income and consumer preferences. The BEA wanted to also reflect the
increased importance of services in consumer purchases, in addition to adding a variety
of new product categories (Bureau of Economic Analysis, 2009).
Personal Consumption Expenditures by product category are now classified into
the following broad categories: (a) Durable Goods, including motor vehicles, household
furnishings and durable household equipment, recreational durable goods and vehicles,
and other durable goods; (b) Non-Durable goods, including food and beverages
purchased to be consumed outside the home, clothing, footwear, gasoline and other
energy purchases; and (c) Services, which include housing and utilities, health care,
transportation services, food services and accommodations, financial services and other
services, including education. In addition to maintaining data on gross expenditures and
quantity indices for its associated product categories, the BEA also maintains Price
Indices for Personal Consumption Expenditures by Type of Product (Bureau of
Economic Analysis, 2009). This series of indices for all product categories will provide
Higher Education and the Cost Disease 52
the dataset from which to compare growth in higher education costs and tuition prices to
other product categories in the economy.
Higher Education Price Index
Most of the discussion focusing on rising college cost and tuition prices has
centered on those increases relative to the consumer price index. However, it has long
been recognized that the CPI may not be an adequate measure of costs in higher
education since the market basket of goods and services utilized by the higher education
community is radically different than those purchased by individual consumers. The
Higher Education Price Index was created to fill this gap. HEPI was first developed by
Kent Halstead beginning in 1961 as an inflation index for a market basket of goods and
services purchased by colleges and universities (Hanson, 1993, p. 46).
The HEPI index measures the change in the relative price for the same bundle of
goods and services over time. The market basket includes the goods and services
purchased by colleges and universities through educational and general expenditures in
the institution's operating budget, excluding expenditures dedicated to research. Items
included in the calculation of the HEPI index include instruction, departmental research,
expenditures for public service, student services, general administration expenditures,
staff benefits, spending on libraries, and physical plant operation and maintenance
(Commonfund Institute, 2004).
As an inflation index, HEPI provides an indictor of the minimum increases
institutions of higher education require to maintain purchasing power over time. Thus,
only an increase in per-unit expenditures above the adjusted inflation rate would reflect
an actual increase in real resources invested in education (Commonfund Institute, 2004,
Higher Education and the Cost Disease 53
p. 1). Seen in this way, the HEPI index provides institutions of higher education a
benchmark from which to assess the minimum increases required to maintain the same
level of goods and services over time (Commonfund Institute, 2005, p. 8).
HEPI measures price levels using a particular year as a reference point. Currently,
1983 is used as a base year (the same base year used by the Consumer Price Index),
which is assigned an index number of 100.00. The indexed value of 1 year is used to
assess the relative changes in prices from the base year. For example, the HEPI index for
the year 2000 was 169.9, meaning that by the year 2000, the average price of goods and
services consumed by colleges and universities had increased 69.9 percent compared to
1983.
The method of calculating the HEPI index has changed over time. Between 1961
and 2001, HEPI was calculated based on price data for over 100 items purchased by
colleges and universities. The components were weighted based on the relative
proportion of expenditures each item represented in an average of college and
universities' operating educational and general budget. For example, personnel costs
comprised 74.8 percent of the total HEPI expenditure index, while contracted services,
supplies and equipment comprised the remaining 25.2 percent of the index weights
(Commonhd Institute, 2004).
Starting in 2002, the HEPI index has been calculated using a regression formula,
based on eight of the original HEPI subcomponents. These include: (a) Faculty Salaries,
(b) Administrative Salaries, (c) Clerical Salaries, (d) Service Employees, (e) Fringe
Benefits, (0 Miscellaneous Services, (g) Supplies and Material, and (h) Utilities. These
eight components equal 79.8 percent of the total weighted average of the all
Higher Education and the Cost Disease 54
subcomponents use in the 1990 calculation. The R-squared value of the regression model
based on the eight subcomponents using the 41 observations based on the original
method of calculating the HEPI index is equal to ,999997809, using the 41 observations
of HEPI index based on the original calculation method. This means that the HEPI values
derived from the regression formula should not deviate from the calculated index by
more than +I- .05 percent (Commonfund Institute, 2004, p. 18).
While the new method for calculating the HEPI index does lose some critical data
since it no longer calculates many of the former HEPI subcomponents, it nevertheless
still provides a useful measure of the inflationary pressures faced by colleges and
universities, and the remaining sub-indices still provide useful data in deconstructing cost
pressures facing institutions of higher education.
Statistical Methods
As described above, the researcher will approach this analysis in two distinct
steps. This will involve utilizing historical time series HEPI and NIPA datasets to
determine changes in spending patterns within higher education, utilizing the Higher
Education Price Index, as well as historical NIPA data to compare higher education cost
and price increases to other national product categories during the same period. Data
from 1961 to 2008 are then compared (the last year data is available for both datasets).
First, the researcher deconstructed the HEPI index to determine the influence of
personnel costs on higher education compared to other HEPI subcomponents. In order to
do this, HEPI and its subcomponents have been re-indexed to a base year of 1961, with
an assigned index value of 100.0 (currently the base year is 1983). Having completed
these data transformations, the various sub-categories can be compared to determine
Higher Education and the Cost Disease 55
which components of institutional spending increased most rapidly during the period
under study, influencing total costs.
This analysis utilizes descriptive statistics, including the HEPI indices themselves,
and an assessment of mean increases over time. Since price indices are designed as an
interval scale, relative distances in index numbers provide powerful insight into cost
dnvers within higher education. This analysis will attempt to answer questions associated
with research sub-questions 1 through 2.
The second phase of this study involves comparing cost and price increases
associated with higher education to national time series data maintained by the Bureau of
Economic Analysis' Personal Consumption Expenditure by Product Category.
Historical time series data also is analyzed for all 69 detail product categories maintained
by the BEA since 1961.
Currently, PCE uses 2005 as a baseline year, with each indexed assigned a value
of 100.0, so each of the PCE price indices has been re-indexed to the study comparison
year of 1961. This provides a comparative context to assess cost and price increases in
higher education since 1961 to price increases in other product categories.
The researcher compared both cost increases associated with the HEPI index as
well as price increases in higher education (as one of the price indices maintained by the
PCE classification) to the other PCE indices. This is particularly important for an industry
such as higher education. Institutions of higher education receive revenue from multiple
sources; changes in relative funding streams may also affect tuition sticker prices, which
are unrelated to either the industry's production function, or the behavioral motivations of
non-profit colleges and universities. Comparing both cost and price indices associated
Higher Education and the Cost Disease 56
with higher education to the price indices associated with other product categories will
also highlight the differences between higher education cost increases based on intrinsic
production costs to price increases which also reflect changes in sources of support.
While descriptive statistics will be analyzed including increases in mean price
among the various product categories, this section will also utilize methods of inferential
analysis. A key premise of cost disease theory lies in the differences between prices
associated with various sectors of the economy. It is anticipated that labor-intensive
industries have higher rates of cost increases than those associated with the
manufacturing sector, and higher education costs and prices should be more reflective of,
and to some extent, mirror these price increases. This should help prove that cost and
price increases are not exclusively based on factors unique to higher education, but are
instead reflective of factors associated with the industry's heavy reliance on labor inputs
to produce its core product.
This requires a comparison of cost and price increases in higher education to test
whether and to what extent differences occur based on industrial sector. The researcher
conducted an analysis of variance comparing both cost and price increases in higher
education to aggregated NIPA indices based on industrial sector. The new NIPA
classification system developed in 2009 facilitates a comparison of data across these
industrial sectors. Personal Consumption Expenditures are classified into three broad
sectors: (a) Durable Goods, (b) Non-Durable Goods, and (c) Services. ANOVA was
conducted to assess the extent to which higher education cost increases (as reflected in
the HEPI index) and higher education price increases (reflected in the PCE category
associated with higher education) compare to increases associated with PCE categories.
Higher Education and the Cost Disease 57
The test is designed to detect differences between two or more groups. This
involved a one factor analysis of variance; the independent variable is industrial sector,
with the dependent variable change in index value during the period under investigation.
In this case, the null hypothesis is that the sample index values associated with the HEPI
index, the Higher Education PCE and the observed indices associated with Durable
Goods, Non-Durable Goods, and Services are equal. ( H O : Price Increases in HEPI = Cost
Increases in Higher Education PCE = Cost Increases Durable Goods PCE = Cost
Increases Non-Durable Goods PCE = Cost Increases Service PCE Services.) The
alternate hypothesis is that differences exist among cost increases among one or more of
the measured indices.
Additionally, ANOVA post-hoc multiple comparison tests were conducted for all
aggregate observations associated with the HEPI Index, Durable and Non-Durable goods,
as well as Services for each year from 1961 through 2008. This was used to assess the
extent to which statistical differences exist among higher education HEPI cost increases,
NIPA higher education price increases, NIPA Durable good increases, NIPA Non-
Durable Good price increases, and increases associated with NIPA-classified service
items.
ANOVA post-hoc multiple comparison tests in which HEPI cost increases, NIPA
higher education price increases, and NIPA Service Good increases are statistically
different (at .05 level of significance) from price increases associated with Durable and
Non-Durable Goods are considered evidence that a cost disease impacts both higher
education in particular and service sector purchases more generally.
Higher Education and the Cost Disease 58
Since a core component of Cost Disease Theory indicates that price increases
associated with service sector industries should outpace those associated with the
manufacturing sector (represented by Durable Good purchases), while Non-Durable
goods, which are generally more labor-intensive than those associated with Durable
Goods but less dependent than services on labor, should also have price increases higher
than those associated with the pure service sector. This test helps discern whether
purchases associated with the manufacturing sector outpace those in other sectors of the
economy. This provides answers to research sub-questions 3 through 5.
The ANOVA provides a quantitative measure to assess whether and to what
extent differences exist between price increases in higher education and broad industrial
sectors as defined by the PCE. This provides a more clinical approach in assessing the
causes for higher education cost and price increases, and perhaps, more fully inform the
current policy debate.
Significance and Limitations of the Study
Signzficance of Study
There is increasing anxiety, confusion, and growing anger over the rate of tuition
price increases which have steadily outpaced increases in the consumer price index.
Many in the political branches of government have been expressing growing frustration
with tuition price increases, and laying the blame at institutions themselves, convinced
that colleges and universities have been increasing prices beyond what is necessary,
based on a misdrected sense of priorities. Defenders of higher education have often
quoted variations of Cost Disease Theory as a way to explain tuition price increases, but
very little work has been conducted to attempt to assess the extent to which price
Higher Education and the Cost Disease 59
increases in higher education can be explained by the existence of a Cost Disease. Studies
analyzing the presence of a cost disease have been conducted for a variety of service
sector industries, however, very few focused on higher education. While a recent study
conducted by Archibald and Feldrnan (2008) attempted to assess whether higher
education has experienced similar cost increases as other service sector industries, their
conclusions were limited by their approach to the data, and although conducted in 2008,
they limited the period under study from 1929 through 1996.
Much of the public policy discussion surrounding public financing of higher
education has centered on the causes of rapidly rising tuition prices. Meanwhile, colleges
and universities have been increasingly relying on adjunct labor as a way to limit costs. A
better understanding of the causes of cost and price increases in higher education may
assist in the public policy discussion. If faster-than-inflation costs are based on factors
idiosyncratic to the higher education industry, then policy solutions centering on greater
regulation and oversight may be appropriate. However, if faster-than-inflation price
increases are similar to other labor intensive industries, and have been exacerbated by a
relative decline in public support, this might suggest that government support should be
increased. Crucial public policy choices may be affected by a better understanding of the
causes for cost and price increases in higher education.
Limitations of the Study
This study has focused on the extent to which tuition price increases in higher
education can be explained by the presence of a cost disease associated with the higher
education production function. The study is limited by the availability of data in
important ways. As originally conceived, the Higher Education Price Index was designed
Higher Education and the Cost Disease 60
as an aggregate measure of cost increases across the whole spectrum of higher education.
It is a broad metric, which encompasses both private and public institutions. However,
there are significant differences in the revenue streams upon which private non-profit and
public institutions depend, which may impact tuition prices independent of underlying
cost structures. Since public institutions are much more dependent upon public sources of
support, they may be more affected by relative changes in h d i n g sources, which may
then become reflected in tuition prices. Similarly, as witnessed during the 2009-10
academic year, changes in the value of stock portfolios can significantly impact
endowment income targeted to operations, which again may impact tuition prices,
unrelated to the costs of providing education.
While recognizing these important differences, a study analyzing both cost drivers
within higher education and comparing aggregate cost and price increases in higher
education to other sectors of the economy provides useful information for current policy
debates. Assessing the extent of a relationship between a cost disease and the entire
higher education sector may assist future policy discussions by creating a better
understanding of cost pressures which may be relatively outside institutional control to
moderate, as opposed to more discretionary spending which for which institutions may be
more accountable for containing.
Similarly, this research has focused on the extent to which a cost disease
influences higher education costs higher education, and may be responsible for rising
higher education costs and tuition prices. This does not necessarily negate research
examining the complex motivations guiding the decision-making of non-profit
institutions. There is a great deal of evidence that the goals of non-profit education
Higher Education and the Cost Disease 61
colleges and universities and competition among institutions for student peer-quality no
doubt contribute to spending pressures. These explanations may not, in fact, be mutually
exclusive, but instead could be mutually reinforcing; pressures on costs and prices arising
from the institutional goals may be compounded by the nature of the costly inputs which
are essential in defining measures of institutional excellence, especially reliance on
relatively scarce labor which may be immune to productivity increases equal to other
non-labor intensive industries. The nature of the competition for excellence may also fuel
added pressures to increase institutional non-need and merit based aid as a way to
leverage student 'quality', seen as vital to institutions as they attempt to maximize the
peer effects of students in the their educational experience.
Future research may attempt to analyze the cost and price behavior of colleges
and universities by incorporating the behavioral aspirations of colleges and universities,
as well as examining intrinsic production costs based on higher education's production
function. A key framework for a more fully holistic model may lie in the work conducted
by Winston and his colleagues at the Williams Project on the Economics of Higher
Education. Winston (1996,2003) has developed a truly comprehensive microeconomic
model of the higher education market which can help explain institutional spending and
tuition pricing behavior. However, this research did not examine the effects of the Cost
Disease on college and university costs, or the effects of burden shifting. Incorporating
both behavioral components of college and university goals with the innate cost pressures
faced by higher education as a labor intensive industry may provide greater insight into
fully explaining cost and price pressures in higher education.
Higher Education and the Cost Disease 62
CHAPTER IV: RESEARCH FINDINGS
Higher Education Price Index: Integrating HEPI Data Series
As described in Chapter IV, a primary piece of evidence for the existence of a
cost disease is the extent to which labor costs have driven overall costs within higher
education. This requires an assessment of the changes in the higher education price index
since its creation in 1961, through 2008, the last year for which complete data exists for
both the Higher Education Price Index the National Income and Product Accounts
dataset.
However, this offers some challenges. The calculation of the HEPI index
underwent a fundamental change after 2001. Initially, the HEPI index was based on Price
Data for over 100 items purchased by colleges and universities. The components were
weighted based on the relative proportion of expenditures for each item represented in an
average of college and universities' operating education and general budget. Personnel
costs comprised 74.8 percent of the total HEPI expenditure, while the contracted services,
supplies and equipment comprised the remaining 25.2 percent of the weighted index
(Commonfund Institute, 2004).
Starting in 2002, the HEPI index has been calculated using a regression formula,
based on eight of the original HEPI components. These include: (a) Faculty Salaries, (b)
Administrative Salaries, (c) Clerical Salaries, (d) Service Employees, (e) Fringe Benefits,
(f) Miscellaneous Services, (g) Supplies and Materials, and (h) Utilities. These eight
components total 79.8 percent of the total weighted average of all the subcomponents
used in earlier calculations. The R-squared value of the regression model using the eight
subcomponents as independent variables for the 41 observations based on the original
Higher Education and the Cost Disease 63
method of calculating the HEPI index is equal to ,999997809. This means that the HEPI
values derived from the regression formula should not deviate from the calculated index
by more than +I- .05 percent (CommonFund, 2004).
A researcher attempting to integrate the two HEPI datasets faces two challenges;
first, unfortunately, most of the detail component analysis conducted by Kent Halstead,
the original developer of the Higher Education Price Index, is no longer available. The
researcher contacted the CornrnonFund Institute to request permission to use any legacy
data which they might make available. The Director of the Institute was eager to provide
as much assistance as possible, and released all legacy data maintained by the institute.
This, however, was minimal. Fortunately, the researcher was able to cull a much more
extensive dataset from prior published reports which has proved extremely useful, and
can serve as a base for conducting systematic analysis of costs since 1961. This includes
cost indices from 1961 through 2001 for Professional Salaries, Non-Professional Salaries,
Fringe Benefits, Cost Indices for Faculty, Graduate Assistants, Executive/Public Service,
AdministrationIInstitutional Services, Library Personnel, and Supply and Equipment
since 1983 (see Appendix B for initial dataset).
As stated above, the second challenge facing a researcher attempting to utilize the
Higher Education Price Index is the change in calculating the index since 2001; thus, the
question becomes how can the data collected between 1961 through 2001 be integrated
with the HEPI index calculated since 2001 based on the eight components? Fortunately,
there is a solution: part of the missing dataset can be reconstructed using a regression
model, while other components can be approximated using the weighting system used in
the initial calculation of the HEPI index. Each of these methods will be addressed in turn.
Higher Education and the Cost Disease 64
The pre-2002 legacy HEPI dataset contained a data series calculating the
aggregate cost of contracted services, one of two primary components used in the
calculation of the entire HEPI series. This is an important measure to determine the
relative costs associated with Personnel and Non-Personnel components of the HEPI
index, and the extent to which each may be considered a cost driver fueling higher
education costs and prices.
Fortunately, there is a way to derive this information based on the information
available. The current data series contains three data elements which account for 78.8 %
of the weight associated with the older Contracted Services, Supplies, and Equipment
data series, including Miscellaneous Services (30.6%), Supplies, and Materials (17.4%),
and Utilities (30.8%). Moreover, information for these three components are available
from 1961 to the present.
To calculate the index weight for Contracted Services, Supplies, and Equipment
from 2002 through 2008, the researcher developed a linear regression model based on the
following components: [Contracted Services, Supplies and Equipment] = [Services] +
[Supplies and Materials] + [Utilities], for the time series associated with the years 1961
through 2001. In the model, [Contracted Services, Supplies and Equipment] is the
dependent variable, while [Services], [Supplies and Materials], and [Utilities] are the
independent variables.
The model was found to be highly significant, with an adjusted r-square value of
,999, and a significance of ,000. Meinwhile, each of the independent variables was also
found to be significant at the ,000 level (see Table 1). Using Beta values associated with
the model, values for Contracted Services, Supplies, and Equipment for the period
Higher Education and the Cost Disease 65
between 2002 through 2008 will be calculated using the following formula: [Contracted
Services, Supplies and Equipment] = -6.391 + ,550 * [Miscellaneous Services] + ,336 *
[Supplies and Materials] + ,164 * [Utilities]. Using this model, the data series associated
with Contracted Supplies, Services, and Equipment can be extended through 2008,
providing another critical piece of data in determining cost drivers within higher
education between 1961 and 2008,
Table 1.
Regression Model Used to Calculate Contracted Services, Supplies and Equipment Variables EnteredlRemoved
Variable Model Variable Entered Removed Method
Utilities, Misc Services, Supplies
1 and Materials Enter a. All requested variables entered b. Dependent Variable: Total Contracted Services
Model Summary
Standard Adjusted Error of the
Model R R-Square R-Square Estimate 1 1.000a ,999 ,999 1.2705
a. Predictors: (Constant), Utilities, Misc Services, Supplies and Materials
ANOVAb Mean
Model Sum of Squares Df Squares F Sig 1 Regression 88555.629 3 29518.543 1.83E+04 .000a
Residual 59.721 37 1.614 Total 88615.35 40
a. Predictors: (Constant), Utilities, Misc Services, Supplies and Materials b: Dependent Variable: Total Contracted Services
Higher Education and the Cost Disease 66
Table 1, Continued.
Regression Model Used to Calculate Contracted Services, Supplies and Equipment Coefficients
Standardized Unstandardized Coefficients Coefficients
Model B Std. Error Beta t Sig 1 (Constant) -6.391 ,673 -9.498 ,000
Misc Services ,550 ,015 .611 37.861 ,000 Supplies and Materials ,336 ,028 ,259 12.169 ,000 Utilities ,164 ,018 ,143 9.284 ,000
Other pieces of the time series associated with Personnel expenditures are missing
and must be reconstructed as well. This poses some additional challenges. Personnel
Compensation is one of the two main components associated with the HEPI index, which
was independently calculated between 1961 and 2001, along with calculations for
Professional and Non-Professional Services. However, the post-2001 regression-based
HEPI index discontinued the calculation of these distinct series, instead using five
subcomponents associated with Personnel Expenditures as part of its regression equation.
These subcomponents include: (a) Fringe Benefits, (b) Faculty Salaries, (c)
Administrative Salaries, (d) Clerical and (e) Service Employees. Aggregate Personnel
Compensation time series index data exists from 1961 to the present, however, faculty
and administrative index data are available beginning in 1981 through 2008, while
distinct index data for Clerical and Service Personnel are only available beginning in
Together, Faculty and Administrative Salaries comprise 87.5 percent of the
weight attributed to Professional Salaries in the pre-regression HEPI index, while Clerical
Higher Education and the Cost Disease 67
and Service salaries comprised 62.3 percent assigned to the Non-Professional Wages
index (see Table 2). Using the relative weights associated with the Pre-regression index
on the available index data, the researcher recalculated the Professional and Non-
Professional Salaries indexes fiom 2002 through 2008 based on the remaining available
data series. For example, in the pre-regression HEPI index associated with Professional
Salaries, Faculty Salaries were allocated a weight of 72.6 percent of the entire
Professional Salaries index, while, Administrative Salaries were assigned a weight of
14.9. Since these are the only time series datasets which were calculated after the 2002
method of calculating the formula, the values associated with faculty and administrative
salaries have been re-weighted based on their relative values in the pre-regression index
to derive the index associated with Professional Salaries for 2002 through 2008. Similar
calculations were developed to derive Non-Professional Services based on the time series
associated with Clerical and Service Personnel between 2002 through 2008. While this
obviously gives a slightly greater weight to faculty and administrative salaries in the
calculation of professional salaries, and the Clerical and Service employee indices in the
calculation of the non-professional salaries index, this should be relatively minimal since
these components account for a significant weighting in the original calculation of the
legacy index. (See Appendix C for enhanced HEPI dataset including Contracted Services,
Supplies and Equipment, Professional and Non-Professional Salaries from 2002-2008.)
Higher Education and the Cost Disease 68
Table 2:
PERSONAL COMPENATION 77.8
Percent Distribution of College and University Current Fund Educational and General Expenditures, Budget FY I983
Weights Associated with the Higher Education Price Index
Professional Salaries Faculty Graduate Assistants Extension and Public Service Adnun and inst Services Library
Non-Professional Salaries Technicians Craftsmen Clerical Students Service Operators and laborers
Fringe Benefits
CONTRACTED SERVICES, SUPPLIES, EQUIPMENT 25.2 Services 30.6
Data Processing 16.4 Communication 16.6 Transportation 11.5 Printing and duplication 7.3 Miscellaneous Services 48.2
100.0
Supplies and materials Equipment Library acquisitions . A
Utilities - 30.8 100.0
Higher Education and the Cost Disease 69
Research Question 1: What are the main cost drivers responsible for driving the Higher Education Price Index?
Table 3 displays the differences between the CPI and the HEPI Index between
1961 and 2008. The table displays annual increases in the CPI and HEPI since 1962, as
well as a standardized CPI and HEPI index, which uses 1983 as a base year set to 100.
As can be seen from the table, the Higher Education Price Index has increased 173.2
percent between 1983 and 2008, while the CPI has increased by a relatively lower rate of
116.7 percent within the same period of time.
Table 3:
Historical Summary of the Consumer Price Indsc and Higher Education Price Index Yearly Percentage Increases, FY 1961 through FY 2008, HEPI Index 1983 - 100
Indexes 1983 = Indexes 1983 =
Year Yearly % 100 Year Yearly % 100 CPI HEPI CPI HEPI CPI HEPI CPI HEPI
1961 30.3 25.6 1985 3.9% 5.7% 107.7 110.8
Higher Education and the Cost Disease 70
Table 3, Continued.
Historical Summary of the Consumer Price Index and Higher Education Price Index Yearly Percentage Increases, FY 1961 through FY 2008, HEPI Index 1983 - 100
Indexes 1983 = Indexes 1983 =
Year Yearly % 100 Year Yearly % 100 CPI HEPI CPI HEPI CPI HEPI CPI HEPI
1977 5.8% 6.4% 59.8 61.5 2001 3.4% 6.0% 178.4 208.7 1978 6.7% 6.8% 63.8 65.7 2002 1.8% 1.9% 181.6 212.7 1979 9.4% 7.3% 69.8 70.5 2003 2.1% 5.1% 185.5 223.5 1980 13.3% 9.9% 79.1 77.5 2004 2.2% 3.7% 189.6 231.7 1981 11.5% 10.7% 88.2 85.8 2005 3.0% 3.9% 195.3 240.8 1982 8.6% 9.4% 95.8 93.9 2006 3.8% 5.1% 202.7 253.1 1983 4.4% 6.5% 100.0 100.0 2007 2.6% 2.8% 208.0 260.3 1984 3.7% 4.8% 103.7 104.8 2008 3.7% 5.0% 215.7 273.2
Source: CommonFund Institute, 2009 HEPI Update
While Table 3 provides tantalizing clues about the extent of price increases in
higher education compared to the consumer price index during the past five decades,
more information is needed to ascertain the nature of the component increases over time.
Table 4 details the nature of price changes for individual components of the Higher
Education Price Index, for which continuous data have been collected sine 1983 through
the present, again compared to the total HEPI and CPI indices. The CPI uses 1983 as a
base year, and this standard is followed in the calculation of the HEPI index. This allows
easy comparisons over time to assess changes in relative price for each of the components
making up the HEPI index.
Higher Education and the Cost Disease 71
Table 4:
Consumer Price Index, Higher Education Price Index, and Major HEPISubcomponents, 1983-2008
Contracted Services, Supplies Year CPI HEPI Personnel Components and Equipment
Supplies Faculty Admin Fringe Misc and Salaries Salaries Beneifts Services Materials Utilities
2008 215.7 273.2 268.1 314.0 380.7 246.4 180.0 252.0 Source: CornmonFund Institute (2004), College and university higher education price index and ComrnonFund Institute (2009) 2009 HEPI Update: CornmonFund Institute. Wilton CT.
Table 4 provides greater clues concerning cost drivers within higher education
since 1983. As before, aggregate prices within higher education have increased 173.2
Higher Education and the Cost Disease 72
percent between 1983 and 2008, while the consumer price index has increased by 115.7
percent within the same period. However, cost increases within higher education have not
been uniform. Between 1983 and 2008, costs associated with faculty salaries, which
exclude benefits, have increased by 168.1 percent during the same period, which is
actually slightly below increases associated with the aggregate HEPI index. Meanwhile,
costs associated with Administrative Salaries have increased at an even faster rate, for an
aggregate of 214 percent during the 25 years displayed in the table. However, cost
increases associated with fringe benefits have far outpaced either of the salary indices,
increasing by 280.7 percent between 1983 and 2008.
In contrast, costs associated with Contracted Services, Supplies, and Equipment,
indices for which continuous data exist between 1983 and 2008 show a much slower
pattern of growth. Miscellaneous services, while increasing faster than the C P by 146.4
percent between 1983 and 2008, grew at a rate much lower than the aggregate HEPI
index (173.2 percent). During the same period, costs associated with supplies and
materials increased by only 80.0 percent, which was even lower than increases in the CPI
during the same period (1 15.7 percent). College and University utility costs have
undergone tremendous fluctuations during the period. Up until 2004, utility costs grew at
a rate even below the CPI. These costs have increased substantially since 2004,
increasing by 152 percent between 1983 and 2008; however, even with rapid increases in
fuel prices since 2004, these costs were below those associated with the aggregate HEPI
index.
Higher Education and the Cost Disease 73
Table 5:
CPI, HEPI, Professional and Non Professional Salaries, Fringe Benefits, Total Personal Compensation and Total contracted services, Supplies and ~&ipme;t FY 1983 to2008.1983 = 100
Year CPI HEPI Personal Compensation
Non Total Professional Profession Fringe Personal Salaries al Salaries Benefits Compensation
Total Contracted Services
100.0 103.0 107.1 109.0 107.4 109.8 112.8 118.7 123.3 126.9 129.4 133.6 135.3 139.9 147.2 151.6 150.8 155.8 172.2 169.2 179.1 187.1 197.8 214.5 216.4
Higher Education and the Cost Disease 74
Source: Commonfund Institute. (2004). College and university higher education price index. Wilton, CT: CommonFund Institute, and Commonfund Institute. (2009). 2009 HEPI Update: Commonfund Institute.; Professional Salaries, 2002-2008 based on weighted average of Faculty and Administrative Salaries; Non-Professional Salaries, 2002-2008 based on weighted average of Clerical and Service Employee Salaries; Total Personal Compensation based on Weighted Average of Professional, Non-Professional and Fringe Benefits; Total Contracted Services based on Regression Formula using Miscellaneous Services, Supplies and Materials, and Utilities using data from 1961-2001
Table 5 provides additional evidence for cost drivers within higher education.
When fringe benefits are included, aggregated costs associated with personnel
compensation increased 185 percent between 1983 and 2008, outpacing increases in the
aggregate HEPI index, which increased by 173.2 percent within the same period of time,
while contracted services increased at the much slower rate of 130.9 percent. Until 2004,
inflation increases associated with Contracted Services actually grew at a rate lower than
the Consumer Price Index, 87.1 percent, compared to the increase in the CPI of 89.6
percent during the same period. Even with rising Contracted Service Costs since 2004,
increases associated with Contracted Services grew moderately faster than the CPI, with
Contracted Services increasing 130.9 percent between 1983 and 2008, while the CPI
increased by 115 percent during the same period. The information displayed in Table 4
indicates that, at least for the time period associated with 1983-2008, personnel costs
were the main cost drivers within higher education.
There has been some concern that the period beginning in 1983 may not be
representative of the nature of cost increases in higher education, especially as related to
salary increases. Due to double-digit inflation in the 1970's, as measured by the CPI,
which sometimes outpaced cost increases in higher education, the increases in salary
during this era have been considered somewhat aberrant, and merely reflective of salary
adjustments making up for lost wages during the prior decade.
Higher Education and the Cost Disease 75
To properly assess higher education cost increases before 1983, this report
includes HEPI and CPI analyses as of 1961, when the HEPI index was created, using the
aforementioned data integration methods. This requires that the HEPI and CPI indices,
using 1983 as a base year set to 100, be re-indexed to the year 1961. (See Appendix D for
re-indexed HEPI dataset, Table 6 displays some of these results.)
Table 6 provides evidence for cost drivers within higher education over a longer
period of time, from 196 1 through 2008. Between 1961 and 2008, the consumer price
index increased 61 1.88 percent, while the Higher Education Price Index increased at a
much faster rate, totaling 997.19 percent during the same period. However, there has
been great variability in increases among the components of the Higher Education Price
Index.
Both Professional and Non-Professional base salaries increased faster than the
inflation as defined by the Consumer Price Index: Professional Salaries increased 859.36
percent between 1961 and 2008, while non-professional salaries increased 735.29 percent
during the same period of time.
However, fringe benefits associated with both professional and non-professional
salaries skyrocketed during the same period, increasing 3,907.37 percent between 1981
and 2008. When fringe benefits are factored into total personnel compensation, personnel
compensation increased at a rate faster than the Higher Education Price Index, for a total
of 1,022.57 percent between 1961 and 2008.
Cost associated with components of Contracted Services, Supplies, and
Equipment in general increased at a much slower rate: Miscellaneous Services increased
648.94 percent between 1961 and 2008, much slower than the general HEPI index and
Higher Education and the Cost Disease 76
only slightly faster than the Consumer Price Index. Costs associated with Supplies and
Materials increased 434.3 1 percent during the same period of time, much slower than
even general inflation measured by the CPI. Between 1961 and 2008, total increases in
utilities rose significantly faster than either the CPI or HEPI, for a total rate of 1,505.10
percent; however most of the faster than HEPI rate increases occurred beginning in 2001;
before that time, increases in utility costs were below those associated with the Higher
Education Price Index. Since 2001, utility costs have been a factor in driving total HEPI
costs faster than the CPI. However, total Contracted Supplies and Equipment, including
fuel costs, increased 784.82 percent between 1961 and 2008, only slightly higher than
increases associated with general inflation as reflected in the Consumer Price Index, and
well below increases associated with the entire HEPI index (967.19 percent).
Higher Education and the Cost Disease 77
Table 6:
Consumer Price Index, Higher Education Price Index, and Major HEPISubcomponents, 1961-2008
Personnel Components Contracted Services, Supplies and Equipment
CPI HEPI Year Index Index
1961 100.00 100.00 1962 100.99 103.52 1963 102.31 107.81 1964 103.63 111.72 1965 104.95 116.41 1966 107.59 121.48 1967 110.56 128.52 1968 114.19 136.33 I969 119.80 144.92 1970 127.06 154.30 1971 133.66 164.45 1972 138.28 173.05 1973 143.89 182.42 1974 156.77 194.92 1975 174.26 212.11 1976 186.47 225.78 1977 197.36 240.23
Non Total Professional Professional Fringe Personal
Salaries Salaries Benefits Compensation 100.00 100.00 100.00 100.00 104.88 102.85 107.37 104.33 110.45 105.34 116.84 109.45 115.33 107.83 129.47 114.57 121.60 110.32 136.84 120.08 128.57 113.17 157.89 127.17 139.02 117.44 176.84 135.04 145.64 123.49 198.95 144.09 155.40 130.25 229.47 154.33 166.20 138.08 260.00 165.75 174.56 148.75 294.74 176.38 181.18 159.79 327.37 185.43 189.20 169.40 365.26 196.06 199.30 180.07 406.32 207.87 210.10 194.31 451.58 221.65 22 1 .25 209.96 503.16 236.22 231.36 224.56 555.79 250.00
Supplies Misc and Services Materials
100.00 100.00 101.82 99.70 103.95 99.70 106.38 100.30 108.81 100.90 110.94 103.28 114.89 105.37 119.15 107.46 124.32 109.25 130.09 112.24 137.08 116.42 145.29 118.81 151.67 122.69 158.66 138.81 172.64 173.13 179.64 181.19 190.27 190.45
~ & a l Contracted
Utilities Services 100.00 100.00 100.64 101.53 100.64 102.68 100.00 104.21 100.00 105.75 100.00 108.05 100.00 110.73 100.64 113.79 101.27 117.62 103.82 122.22 114.65 130.65 122.29 137.93 128.66 144.06 157.96 158.62 202.55 185.82 219.11 196.55 257.96 213.41
Higher Education and the Cost Disease 78
Table 6, Continued.
Consumer Price Index, Higher Education Price Index, and Major HEPZ Subcomponents, 1961-2008
Personnel Components Contracted Services, Supplies and Equipment
Non Total CPI HEPI Professional Professional Fringe Personal
Year Index Index Salaries Salaries Benefits Compensation 1978 210.56 256.64 243.55 242.35 614.74 266.14 1979 230.36 275.39 258.19 261.21 678.95 285.04 1980 261.06 302.73 276.66 285.41 764.21 308.66 1981 291.09 335.16 300.70 312.10 861.05 337.80 1982 316.17 366.80 326.48 336.65 963.16 368.11 1983 330.03 390.63 348.43 355.87 1,052.63 393.70 1984 342.24 409.38 364.81 374.02 1,140.00 414.96 1985 355.45 432.81 388.15 388.61 1,238.95 440.94 1986 365.68 454.30 41 1.85 401.42 1,344.21 467.72 1987 373.93 472.27 435.54 413.88 1,446.32 493.70 1988 389.44 492.97 456.10 429.18 1,549.47 518.50 1989 407.59 518.75 483.62 445.91 1,671.58 549.61 1990 427.06 550.00 514.29 463.70 1,804.21 583.86 1991 450.17 578.91 542.16 481.85 1,940.00 616.14 1992 464.69 599.61 560.28 498.93 2,045.26 639.37 1993 479.21 616.80 574.91 513.17 2,150.53 659.84 1994 491.09 637.89 593.38 527.40 2,248.42 682.28 1995 505.61 656.64 613.59 542.70 2,330.53 705.12
Supplies Total Misc and Contracted Services Materials Utilities Services
202.43 198.81 292.36 230.65 216.41 214.03 320.38 249.43 234.04 252.54 408.28 287.36 258.97 285.37 507.64 329.12 286.32 299.70 588.54 363.60 303.95 298.51 636.94 383.14 318.84 297.61 652.87 394.64 336.78 307.46 670.70 410.34 349.85 306.27 656.69 417.62 363.83 295.52 579.62 41 1.49 373.86 301.79 558.60 420.69 391.49 323.28 543.3 1 432.18 407.29 341.19 573.89 454.79 424.92 347.46 588.54 472.41 442.86 343.88 594.27 486.21 454.41 337.91 603.18 495.79 470.52 341.19 628.66 511.88 480.24 345.37 616.56 518.39
Higher Education and the Cost Disease
Table 6, Continued.
Consumer Price Index, Higher Education Price Index, and Major HEPZSubcomponents, 1961-2008
Personnel Components Contracted Sewices, Supplies and Equipment
Non Total Supplies Total CPI HEPI Professional Professional Fringe Personal Misc and Contracted
Year Index Index Salaries Salaries Benefits Compensation Services Materials Utilities Services 1996 519.47 675.78 633.10 559.79 2,363.16 724.80 497.87 388.36 594.27 536.02 1997 534.32 696.88 652.26 576.87 2,386.32 744.09 508.51 383.88 675.80 563.98 1998 543.89 721.48 674.22 597.86 2,491.58 770.87 525.23 376.72 707.64 580.84 1999 553.14 738.67 699.30 619.57 2,517.89 795.28 537.99 367.76 640.13 577.78 2000 569.31 769.14 726.13 641.99 2,680.00 829.92 555.93 367.46 668.15 596.93 2001 588.78 815.23 751.92 668.68 2,754.74 858.66 607.29 393.43 1,082.17 659.77 2002 599.34 830.86 784.09 707.22 2,916.84 900.31 625.53 382.69 752.23 648.44 2003 612.21 873.05 816.35 725.41 3,076.84 937.53 636.78 394.63 1,003.82 686.21 2004 625.74 905.08 833.29 743.30 3,292.63 967.08 657.75 404.78 1,123.57 716.94 2005 644.55 940.63 858.43 761.97 3,444.21 998.72 676.90 434.33 1,275.16 757.91 2006 668.98 988.67 888.24 780.67 3,617.89 1,035.14 695.44 471.94 1,628.66 821.86 2007 686.47 1,016.80 922.24 809.70 3,797.89 1,077.17 724.32 493.43 1,405.10 829.09 2008 711.88 1,067.19 959.36 835.29 4,007.37 1,122.57 748.94 537.31 1,605.10 884.82
Source: Commonfund Institute. (2004). College and university higher education price index. Wilton, CT: CommonFund Institute, and Commonfund Institute. (2009). 2009 HEPI Update: Commonfund Institute.; Professional Salaries, 2002-2008 based on weighted average of Faculty and Administrative Salaries; Non-Professional Salaries, 2002-2008 based on weighted average of Clerical and Service Employee Salaries; Total Personal Compensation based on Weighted Average of Professional, Non- Professional and Fringe Benefits; Total Contracted Services based on Regression Formula using Miscellaneous Services, Supplies and Materials, and Utilities using data from 1961-2001. Reweighted using 1961 as base year.
Higher Education and the Cost Disease 80
Research Question 2: To what extent are labor costs driving overall costs within higher education?
Table 7 displays the aggregates for Total Personnel Compensation and Contracted
Supplies and Equipment, along with the HEPI index between 1961 and 2008. When
contrasted with Contracted Supplies and Equipment, it seems clear that total personnel
costs, which include fringe benefits, have been the main cost driver propelling total costs
within higher education (1,022.57 percent increase between 1961 and 2008), while Total
Contracted Supplies and Equipment increased well below the aggregate HEPI Index
(784.82 percent between 1961 and 2008). This provides important evidence that higher
education is affected by a cost disease, revealing that personnel costs are a primary dnver
of costs within higher education. However, in order to identify the existence of a cost
disease, this research must also attempt to ascertain the extent to which higher education
cost and price increases are similar to other service sector industries, and the extent to
which price increases associated with services in general are similar or different to those
associated with the manufacturing sector. These issues will be addressed in the next sub
question.
Higher Education and the Cost Disease 81
Table 7.
Consumer Price Index, Total Personal Compensation and Contracted Services, 1961-2008
Total CPI HEPI Total Personal Contracted
Year Index Index Compensation Services
Higher Education and the Cost Disease 82
Table 7, Continued.
Consumer Price Index, Total Personal Compensation and Contracted Services, 1961-2008
Total CPI HEPI Total Personal Contracted
Year Index Index Compensation Services 1996 519.47 675.78 724.80 536.02 1997 534.32 696.88 744.09 563.98 1998 543.89 721.48 770.87 580.84 1999 553.14 738.67 795.28 577.78 2000 569.31 769.14 829.92 596.93 2001 588.78 815.23 858.66 659.77 2002 599.34 830.86 900.31 648.44 2003 612.21 873.05 937.53 686.21 2004 625.74 905.08 967.08 716.94 2005 644.55 940.63 998.72 757.91 2006 668.98 988.67 1,035.14 821.86 2007 686.47 1,016.80 1,077.17 829.09 2008 711.88 1,067.19 1,122.57 884.82
Source: CommonFund Institute (2004) College and university higher education price index, Wilton, CT; CommonFund Institute, and CommonFund Institute (2009).2009 HEIP Update, CommonFund Institute Personal Compensation, 2002-2008 based on weighted average of Professional and Non-Professional Salaries and Fringe Benefits; Total Contracted Senices based on Regression Formula using Miscellaneous Services, Supplies and Materials and Utilities, using data from 1991-2001. Reindexed using 1961 as base year.
Research Question 3: To what extent can a cost disease explain rapidly rising costs and tuition sticker prices?
The data analyzed to answer questions 1 and 2 indicate that personnel costs are
largely responsible for fueling total costs within higher education. While providing
credence to the theory that higher education is affected by a cost disease, this alone is
insufficient evidence to definitely conclude a cost disease impacts higher education. Cost
Disease theory postulates that labor-intensive industries should experience faster-than-
inflation cost increases compared to those industries which can successfully leverage
Higher Education and the Cost Disease 83
technology to increase productivity. Fully testing the extent to which higher education is
impacted by a cost disease requires a comparison of price increases in higher education
over time to determine whether price increases in higher education are similar to price
increases associated with other labor intensive industries, and also whether there are
differences between labor-intensive industries and those associated with the
manufacturing sector. However, since prices in higher education may fluctuate based on
shifts in support among the relative revenue streams on which higher education relies,
this research also included an assessment of the extent to which changes in the higher
education price index (representing intrinsic higher education production costs) are
similar or different to increases associated with higher education prices, along with price
increases associated with the manufacturing and labor-intensive industries.
The researcher utilized data associated with the National Income and Product
Accounts Personal Consumption Expenditures Index as well as the aggregated HEPI
index between 1961 and 2008 to analyze price increases among all Personal
Consumption Expenditures, Durable Goods, Non-Durable Goods, Services, Higher
Education Prices, and Higher Education Costs. The research sought to determine the
extent to which price increases among these categories of goods and services were
similar to or different from one another using ANOVA post-hoc multiple comparison
tests. However, the NIPA dataset uses a base year of 2005, for which all indices are set to
100 (see Appendix E). While this research sought to analyze price increases since 1961.
The NIPA dataset were transformed to set 1961 as the base year, setting all indices to 100
for the year 1961 (see Appendix F). Having recalibrated all indices using 1961 as a base
Higher Education and the Cost Disease 84
year, Appendix G displays the research database which was used to conduct the ANOVA
analysis.
A one-way ANOVA analysis was conducted at a .05 level of significance, with
six levels. Index Value was defined as the dependent variable, against the following
levels: (a) All PCE Goods and Services, (b) Durable Goods, (c) Non-Durable Goods, (d)
Services, (e) Higher Education Prices, and (f) Higher Education Costs.
The null hypothesis is that the sample index values associated with the HEPI
index, the Higher Education PCE and the observed indices associated aggregate PCE
Goods and Services, Durable Goods, Non-Durable Goods, and Services are equal. (HO:
Price Increases in HEPI = Cost Increases in Higher Education PCE = Cost Increases
Aggregate PCE Goods and Services = Cost Increases Durable Goods PCE = Cost
Increases Non-Durable Goods PCE = Cost Increases Service PCE Services.) The
alternate hypothesis is that differences exist among cost increases among one or more of
the measured indices.
ANOVA tests whether the assumption of equal variance in the dependent variable
is true by partitioning the variance into two components: the variance of the scores within
the six groups under study, and the variance between the group means and the total
group. Variances within groups are attributed to random fluctuations. However,
variations between groups reflect both random fluctuations as well as variations based on
distinct group behavior. If the null hypothesis is true, we would expect to see very little
difference between the variance within groups and the variance between the groups,
indicating that the observations came from similar distributed populations.
Higher Education and the Cost Disease 85
The initial ANOVA test determines the ratio of the between group variation to the
within group variation; if the null hypothesis is true, the ratio of the between group
variance to the within group variance would approximately equal to one. If the null
hypothesis is false, the ratio of the between group variance and the within group variance
should be greater than 1. The extent of the difference, based on the selected level of
significance of .O5, will determine whether the null hypothesis is accepted or rejected.
Table 8 reports the results from the ANOVA test statistic; the F-ratio measures
the ratio between the between group variance by the within group variance. The
significance level associated with the F statistic of ,000 is well below the pre-determined
.05 level of significance which would indicate that the null hypothesis is false, indicating
that the probability that the observed difference in the ratio of the between group and
within group variance would have occurred by chance if the null hypothesis is true is less
than .05; in fact, the probability of obtaining an F-ratio of 19.854 with 5 degrees of
freedom if the null hypothesis is true is less than 1 chance in 1000. This suggests a high
degree of confidence that the null hypothesis, assuming that Price Increases in HEPI =
Cost Increases in Higher Education PCE = Cost Increases Aggregate PCE Goods and
Services = Cost Increases Durable Goods PCE = Cost Increases Non-Durable Goods
PCE = Cost Increases Service PCE Services, can be rejected, while the alternate
hypothesis can be accepted. This indxates that price increases associated with at least one
of the price indices tested is significantly different from some of the others.
Higher Education and the Cost Disease 86
Table 8.
ANOVA Analysis: All PCE Goods and Services, Durable Goods, Non-Durable Goods, Services, Higher Education Prices, and Higher Education Costs
Sum of Mean Index Squares Df Square F Sig. Between Groups 1 .05 1E+07 5 210193.79 19.854 ,000 Within Groups 2.986E+07 282 105872.56 Total 4.037E+07 287
However, an additional challenge remains: while the F-statistic provided
information that at least one of the tested indices differs from others, it did not provide
details on which price measures of the six investigated differ. This requires running an
additional corollary test - the Post-Hoc Multiple Comparison Test to determine which of
the six indices included in the ANOVA analysis differ from one another.
The Tukey Post-Hoc Multiple Comparison Test assesses all two-way (painvise)
comparisons based on the pre-determined significance level (for this analysis, determined
to be .05). In this case, the null hypothesis tests 15 individual painvise comparisons; each
of these painvise comparisons is analyzed below.
1) Price Increases All PCE Goods and Services = Price Increases PCE Durable Goods;
2) Price Increases All PCE Goods and Services = Price Increases PCE Non-Durable Goods;
3) Price Increases All PCE Goods and Services = Price Increases PCE Services;
4) Price Increases All PCE Goods and Services = Higher Education Price Increases;
5) Price Increases All PCE Goods and Services =Higher Education Cost Increases;
6) Price Increases PCE Durable Goods = Price Increases PCE Non-Durable Goods;
7) Price Increases PCE Durable Goods =Price Increases PCE Services;
8) Price Increases PCE Durable Goods = Higher Education Price Increases;
Higher Education and the Cost Disease 87
9) Price Increases PCE Durable Goods =Higher Education Cost Increases;
10) Price Increases PCE Non-Durable Goods =Price Increases PCE Services;
11) Price Increases PCE Non-Durable Goods =Higher Education Price Increases;
12) Price Increases PCE Non-Durable Goods = Higher Education Cost Increases;
13) Price Increases PCE Services =Higher Education Price Increases;
14) Price Increases PCE Services =Higher Education Cost Increases;
15) Higher Education Cost Increases = Higher Education Price Increases;
The Tukey Post-Hoc Multiple Comparison Test uses the Q-Statistic to determine
whether the differences between painvise comparisons differ so much that the null
hypothesis should be rejected. The Q-statistic calculates the differences between group
means; if the mean pairwise difference is high enough, the null hypothesis assuming the
two values associated with the painvise value can be rejected, otherwise it is not rejected
Table 9.
Results of Post Hoc Multiple Comparison Test Index Tukey HSD Multiple Comparisons 95% Confidence Interval
Mean Difference Lower Upper
(I) Item Code (J) Item Code (I- J) Std. Error Sig Bound Bound
All PCE Goods and Services Durable Goods 60.579313 6.6418E+01 ,943 -130.00216 251.16078
Non-Durable Goods -38.093133 6.6418E+Ol ,993 -228.67460 152.48834 Services -137.471487 6.6418E+Ol .306 -328.05296 53.10998
Higher Education - Prices 517.327675* 6.6418E+01 ,000 -707.90915 -326.7462
Higher Education - Costs 219.866161* 6.6418E+Ol ,013 -410.44763 -29.28469
All PCE Goods Durable Goods and Services -60.579313 6.6418E+01 ,943 251 .I6078 -130.00216
Higher Education and the Cost Disease 88
Table 9, Continued.
Results of Post Hoc Multiple Comparison Test Index Tukey HSD Multiple Comparisons 95% Confidence Interval
Non-Durable Goods -98.672446 6.6418E+Ol ,674 -289.25392 91.90903
Services 198.050800* 6.6418E+01 ,036 -388.63227 -7.46933
Higher Education - Prices 577.906988* 6.6418E+Ol .000 -768.48846 -387.32552
Higher Education - Costs 280.445474* 6.6418E+01 ,000 -471.02695 -89.86400
Non-Durable All PCE Goods Goods and Services 38.093133 6.6418E+Ol ,993 -152.48834 228.67460
Durable Goods 98.672446 6.6418E+Ol ,674 -91.90903 289.25392 Services -99.378354 6.6418E+Ol ,667 -289.95983 91.20312
Higher Education - Prices 479.234542* 6.6418E+01 ,000 -669.81601 -288.65307
Higher Education Costs -181.773028 6.6418E+Ol ,071 -372.35450 8.80844
All PCE Goods Services and Services 137.471487 6.6418E+Ol ,306 -53.10998 328.05296
Durable Goods 198.050800* 6.6418E+Ol ,036 7.46933 388.63227 Non-Durable Goods. 99.378354 6.6418E+Ol ,667 -91.20312 289.95983
Higher Education Prices 379.856188* 6.6418E+Ol ,000 -570.43766 -189.27472
Higher Education Costs -82.394674 6.6418E+Ol ,816 -272.97615 108.18680
Higher Education All PCE Goods Prices and Services 219.866161* 6.6418E+01 ,000 326.74620 707.90915
Durable Goods 577.906988* 6.6418E+Ol ,000 387.32552 768.48846 Non-Durable Goods 479.234542* 6.6418E+01 ,000 288.65307 669.81601 Services 379.856188* 6.6418E+Ol ,000 189.27472 570.43766
Higher Education Costs 297.461515 6.6418E+01 ,000 106.88004 488.04299
Higher Education All PCE Goods Costs and Services 219.866161* 6.6418E+Ol ,013 29.28469 410.44763
Durable Goods 280.445474* 6.6418E+Ol ,000 89.86400 471.02695
Higher Education and the Cost Disease 89
Table 9, Continued.
Results of Post Hoc Multiple Comparison Test Index Tukey HSD Multiple Comparisons 95% Confidence Interval
Non-Durable Goods -181.773028 6.6418E+Ol ,071 -8.80844 372.35450 Services 82.394674 6.6418E+01 ,816 -108.18680 272.97615
Higher Education - Prices 297.461515* 6.6418E+Ol ,000 -488.04299 -106.88004
* The mean difference is significant at the .05 level.
Table 9 details the pairwise comparisons analyzed using the Tukey Post-Hoc
Comparison Test, and provides key results to assess first, the extent to which a cost
disease can explain higher education cost and prices, as well as whether differences exist
in the rate of price increases between aggregate PCE Product and Services, Durable
Goods, Non-Durable Goods, and Services. The results among these components provide
additional information on the extent to which price increases in service sector purchases
differ from those associated with durable and non-durable goods, and ultimately, whether
a general cost disease impacts the service sector as well. As indicated, 15 distinct
painvise comparisons were conducted using the Tukey Post-Hoc Comparison Test. Each
of these comparisons will be assessed in detail.
Price increases all PCE goods and services =price increases PCE durable goods.
The mean difference between price increases associated with PCE Goods and
Services and PCE Durable goods is 60.579; this is associated with a significance level of
,943. This is well above the .05 level of significance under which it was determined the
null hypothesis should be rejected. Since the chance of obtaining an approximate mean
difference through chance alone is so high (over 94 percent), the null hypothesis
assuming these come from the same distribution is not rejected. This suggests price
Higher Education and the Cost Disease 90
increases associated with All PCE Goods and Services and PCE Durable Goods are not
significantly different.
Price Increases all PCE goods and services =price increases PCE non-durable goods.
The mean difference between price increases associated with PCE Goods and
Senices and PCE Non-Durable goods is -38..93; this is associated with a significance
level of ,993. This is well above the .05 level of significance under which it was
determined the null hypothesis should be rejected. Since the chance of obtaining an
approximate mean difference through chance alone is so high (in this case over 99
percent), the null hypothesis assuming these come from the same distribution is not
rejected. This suggests that price increases associated with All PCE Goods and Services
and PCE Non-Durable Goods are not significantly different.
price increases all PCE goods and services =price increases PCE services.
The mean difference between price increases associated with PCE Goods and
Services and PCE Services -137.471; this is associated with a significance level of ,306.
Since the chance of obtaining an approximate mean difference based on chance alone is
still extremely high (30.6 percent), the null hypothesis assuming these come from the
same distribution is not rejected. This suggests that price increases associated with All
PCE Goods and Services and those Associated with PCE Services are not significantly
different.
Higher Education and the Cost Disease 91
price increases all PCE goods and services = higher education price increases.
The mean difference between price increases associated with PCE Goods and
Services and Higher Education Price Increases is -517.327. This is associated with a
significance level of ,000. Since the chance of obtaining the same mean difference
through chance alone is extremely low (less than 1 chance in IOOO), the null hypothesis
assuming these come kom the same distribution can be rejected. This suggests that price
increases associated with Higher Education are significantly greater than general price
increases associated with all PCE Goods and Services.
price increases all PCE goods and services = higher education cost increases.
The mean difference between price increases associated with PCE Goods and
Services and Higher Education Cost Increases is -219.867. This is associated with a
significance level of .Ol3. Since the chance of obtaining the same mean difference
through chance alone is extremely low (13 out of 1000), the null hypothesis assuming
these come from the same distribution can be rejected. This suggests that cost increases
associated with Higher Education are significantly greater than general price increases
associated with all PCE Goods and Services.
price increases PCE durable goods =price increases PCE non-durable goods.
The mean difference between price increases associated with PCE Durable Goods
and Non-Durable Goods is - 98.672. This is associated with a significance level of ,674.
This is well above the .05 level of significance under which it was determined the null
hypothesis should be rejected. Since the chance of obtaining an approximate mean
difference through chance alone is extremely high (67.4 percent), the null hypothesis
Higher Education and the Cost Disease 92
assuming these come from the same distribution is not rejected. This suggests that price
increases associated with PCE Durable Goods and those associated with PCE Non-
Durable Goods are not significantly different.
price increases PCE durable goods =price increases PCE services.
The mean difference between price increases associated with PCE Durable Goods
and PCE Services is -.198.051. This is associated with a significance level of ,036. This is
below the .05 level of significance under which it was determined the null hypothesis
should be rejected. Since the chance of obtaining the same mean difference through
chance alone is extremely low (3.6 percent), the null hypothesis assuming these come
from the same distribution can be rejected. This suggests that price increases associated
with PCE Services are significantly higher than those associated with PCE Durable
Goods.
price increases PCE durable goods = higher education price increases,
The mean difference between price increases associated with PCE Durable Goods
and Higher Education Price Increases is -577.907. This is associated with a significance
level of ,000. This is well below the .05 level of significance under which it was
determined the null hypothesis should be rejected. Since the chance of obtaining the same
mean difference through chance alone is extremely low (less than 1 chance in 1000), the
null hypothesis assuming these come from the same hstribution can be rejected. This
suggests that price increases associated with Higher Education are significantly greater
than price increases associated with PCE Durable Goods.
Higher Education and the Cost Disease 93
price increases PCE durable goods = higher education cost increases.
The mean difference between price increases associated with PCE Durable Goods
and Higher Education Cost Increases is -280.445. This is associated with a significance
level of ,000. This is well below the .05 level of significance under which it was
determined the null hypothesis should be rejected. Since the chance of obtaining the same
mean difference through chance alone is extremely low (less than 1 chance in 1000), the
null hypothesis assuming these come from the same distribution can be rejected. This
suggests that cost increases associated with Higher Education are significantly greater
than price increases associated with PCE Durable Goods.
price increases PCE non-durable goods =price increases PCE services.
The mean difference between price increases associated with PCE Non-Durable
Goods and PCE Services -99.379. This is associated with a significance level of ,667.
This is well above the .05 level of significance under which it was determined the null
hypothesis should be rejected. Since the chance of obtaining a similar mean difference
through chance alone is so high (over 67 percent), the null hypothesis assuming these
come from the same distribution is not rejected. This suggests that price increases
associated with PCE Non-Durable Goods and PCE Services are not significantly
different.
price increases PCE non-durable goods = higher education price increases
The mean difference between price increases associated with PCE Non-Durable
Goods and Higher Education Price Increases is -479.234. This is associated with a
significance level of ,000. This is well below the .05 level of significance under which it
Higher Education and the Cost Disease 94
was determined the null hypothesis should be rejected. Since the chance of obtaining the
same mean difference through chance alone is extremely low (less than 1 chance in
1000), the null hypothesis assuming these come from the same distribution can be
rejected. This suggests that price increases associated with Higher Education are
significantly higher than even those associated PCE Non-Durable Goods.
price increases PCE non-durable goods = higher education cost increases.
The mean difference between price increases associated with PCE Non-Durable
Goods and Higher Education Cost Increases is -181.773. This is associated with a
significance level of ,306. This is well above the .05 level of significance under which it
was determined the null hypothesis should be rejected. Since the chance of obtaining an
approximate mean difference through chance alone is extremely high (nearly 31 percent),
the null hypothesis assuming these come from the same distribution is not rejected. This
suggests that price increases associated with PCE Non-Durable Goods and those
associated with Higher Education Costs are not significantly different.
price increases PCE services = higher education price increases.
The mean difference between price increases associated with PCE Services and
Higher Education Price Increases is -379.856. This is associated with a significance level
of ,000. This is well below the .05 level of significance under which it was determined
the null hypothesis should be rejected. Since the chance of obtaining the same mean
difference through chance alone is extremely low (less than 1 chance in 1000), the null
hypothesis assuming these come from the same distribution can be rejected. This
Higher Education and the Cost Disease 95
suggests that price increases associated with Higher Education are significantly higher
than even those associated PCE Services.
price increases PCE services = higher education cost increases.
The mean difference between price increases associated with PCE Services and
Higher Education Cost Increases is -82.395. This is associated with a significance level
of ,816. This is well above the .05 level of significance under which it was determined
the null hypothesis should be rejected. Since the chance of obtaining a similar mean
difference through chance alone is so high (nearly 82 percent), the null hypothesis
assuming these come from the same distribution is not rejected. This suggests that price
increases associated with PCE Services and Higher Education Cost Increases are not
significantly different.
higher education cost increases = higher education price increases.
The mean difference between price increases associated with PCE Services and
Higher Education Price Increases is -297.46. This is associated with a significance level
of ,000. This is well below the .05 level of significance under which it was determined
the null hypothesis should be rejected. Since the chance of obtaining the same mean
difference through chance alone is extremely low (less than 1 chance in 1000), the null
hypothesis assuming these come from the same distribution can be rejected. This
suggests that price increases associated with Higher Education are significantly higher
than higher education cost increases.
Analysis of the post-hoc comparison tests provides some mixed answers to question
3: To what extent can a cost disease explain rapidly rising tuition sticker prices? Three of
Higher Education and the Cost Disease 96
the post-hoc comparison tests provide information to answer t h s question: (a) the
relationship between PCE Services to Higher Education Prices (Post-hoc analysis
question (b) the relationship between PCE Services to Higher Education Costs, and (c)
The relationship between higher education costs and higher education price increases.
The theory of a cost disease would suggest that while the prices of services should be
higher than those associated with the manufacturing sector, the cost and prices of services
should be comparable to one another. It is here that the analysis provides some mixed
results.
A comparison of higher education cost increases to those associated with the
service sector reveals no significant differences in the rate of increase between these two
categories. The mean difference of -.82.39 was associated with a significance level of
,816. The chance of obtaining similar results based on chance alone is so high that the
null hypothesis cannot be rejected. This does suggest that there is no statistical difference
between higher education cost increases and price increases associated with PCE
Services. This would lend support to the idea that a cost disease impacts higher
education.
However, the mean difference between price general PCE Services and higher
education price increases was -379.856, which was associated with a significance level of
,000. The chance of obtaining a similar result based on chance alone is less than 1 in
10,000, leading to a rejection of the null hypothesis that both came from similar
distributions. This supports the idea that higher education prices have been increasing at a
rate faster than general PCE Services.
Higher Education and the Cost Disease 97
Similarly, a comparison of higher education costs and higher education prices
yields a mean difference of -297.461, also significant at the ,000 level of significance.
This suggests that higher education prices have been rising much faster than higher
education costs.
The decision to analyze both higher education cost as well as price increases
stems from the impact which shifting revenue streams may have upon higher education
prices, independent of cost pressures facing higher education; this is the potential
problem associated with burden shifting, discussed earlier. If non-tuition revenue streams
fail to keep pace with costs, tuition revenue may be used to fill the gap. This may help to
explain the different results when analyzing higher education cost and price increases
compared to general services represented by PCE Services. The results indicate that
higher education costs have not been rising significantly faster than prices associated with
the service sector, which would indicate that a cost disease impacts higher education,
whose impact does not appreciably differ from those associated with other services.
However, higher education prices have been rising significantly faster than the service
sector, and indeed, significantly faster than higher education costs. The possible reasons
for this will be explored more fully in Chapter 5.
Research Question 4: Are there similarities between increases in higher education costs and tuition sticker prices and prices in other labor intensive industries?
The research findings described above provide tantalizing evidence concerning the
relation of higher education cost and price increase relative to other service sector
consumption categories. A comparison of Services and higher education costs indicate
Higher Education and the Cost Disease 98
comparable rates of increases. The mean difference between the two data series was -
82.395, which was associated with a ,816 level of significance; the chance of obtaining a
similar mean difference due to chance alone is so high that the null hypothesis could not
be rejected. This suggests that higher education cost increases are similar to price
increases associated with a composite of all Services maintained in the National Income
and Product Accounts Personal Consumption Expenditures Index.
While higher education cost increases are similar to those of Services, price increases
associated with higher education far outpaced price increases in the PCE Service index.
The mean difference between price increases associated with aggregate Services in the
Personal Consumption Expenditure Index and Higher Education Price Increases is
- 379.856, which was associated with a significance level of ,000. The chances of
obtaining similar results based on chance are so low (less than 1 in 10,000) that the null
hypothesis, assuming these indices came from the same distribution, could be rejected.
This indicates that higher education costs are rising much faster than prices associated
with all Services in the PCE Index.
Even a comparison of the rate of cost increases in higher education to aggregate price
increases reveals that higher education prices are rising much faster than higher education
costs. The mean difference between price increases associated with PCE Services and
Higher Education Price increases is -297.46. This is associated with a significance level
of ,000, which again indicates a rare occurrence with very little probability the outcomes
were random.
A key principle of Cost Disease theory suggests that since it harder to leverage labor
saving technology in the service sector compared to the manufacturing sector, the
Higher Education and the Cost Disease 99
intrinsic production costs associated with services will rise more rapidly than those
associated with the manufacturing sector, which will also be reflected in the relatively
faster rate of price increases associated with service sector. While higher education costs
have increased at a rate comparable to other services, higher education prices have far
outpaced increases in services, as well as increases in higher education prices. The
dissonance witnessed in the behavior between higher education costs and prices, in
relation to one another as well as the service sector will be explored more fully in Chapter
v.
Research Question 5: Are there differences between price increases in labor intensive industries compared to those associated with the manufacturing sector?
Answers to this research question are based on the relationship between aggregate
Services reflected in the PCE Services Index, PCE Non-Durable Goods, and PCE
Durable Goods, and would also help discover whether a cost disease impacts the service
sector. According to cost disease theory, due to the inability to leverage labor-saving
technology in the service sector as compared to the manufacturing sector, prices should
rise faster in services relative to manufacturing. As far as the categories of products and
services are concerned, this would mean that researchers would expect to see price
increases in services outpace those associated with durable goods, the primary product of
the manufacturing sector.
There does indeed seem to be differences in the rate of price increases between
Durable Goods and Services associated with the Personal Consumption Expenditure
Index. The mean difference between price increases associated with Durable Goods and
Services was -.198.051; this is associated with a significance level of ,036, which is
Higher Education and the Cost Disease 100
below the .05 level of sigmficance under which it was determined the null hypothesis
should be rejected. The rate of price increases associated with services is significantly
higher than those associated with durable goods, primary evidence that a cost disease
does in fact impact Service-related jobs.
Theory is less clear on the precise relationship between prices in the Services and
those associated with Non-Durable Goods. Non-Durable Goods include purchases on
items such as food and alcohol purchased for the home, clothing, personal care products,
and magazine and newspapers. In principle, researchers should expect to find that the
more labor intensive the task, the faster price increases associated with the particular
product category. However, unlike durable goods which are heavily associated with the
manufacturing sector, Non-Durable goods encompass a strong labor component.
The mean difference between price increases associated with PCE Durable Goods
and Non-Durable Goods is - 98.672, whlch is associated with a significance level of
,674. Since obtaining results equal to this due to chance alone would be extremely
common, the null hypothesis was not rejected. While the mean difference indicates that
durable good prices were slightly lower than those associated with non-durable goods,
the difference was simply not significant. There seems to be very little difference in price
increases between durable and non-durable goods.
Similarly, there was no statistical difference in the rate of price increases between
non-durable goods and services. The mean difference was -99.378, which was associated
with a significance level of ,667. While suggesting that services increased at a faster rate
than non-durable goods, the chance that the results were due to chance are so high the
Higher Education and the Cost Disease 101
null-hypothesis was rejected. There is no significant difference between the rate of price
increases associated with non-durable goods and services.
In conclusion, the results discussed in this section do seem to indicate the
presence of a cost disease associated with the service sector; a primary postulate of cost
disease theory states that, since labor is incidental to the production of goods and
services, and central to the output of service sector jobs, costs associated with services
will rise faster than those associated with the manufacturing sector. The research results
analyzed in this chapter strongly support this theory.
Higher Education and the Cost Disease 102
CHAPTER V: CONCLUSION
The Cost Disease and Higher Education: Evidence from Research Findings
The research conducted for this study has revealed compelling evidence that a
cost disease impacts higher education. Determining whether and to what extent higher
education is affected by a cost disease necessitated two complementary methods of
inquiry, involving an analysis of cost pressures within higher education as well as a
comparative analysis of higher education cost and prices against other industries and
sectors.
First, components of the higher education price index were deconstructed to
isolate the relative impact that personnel and non-personnel related expenditures have
had on the trajectory of higher education costs from 1961 through 2008. Following this,
higher education cost and prices were compared to increases associated with other
aggregate sectors of the economy, including Durable Goods, Non-Durable Goods, and
Services.
The analysis of higher education costs between 1961 and 2008 revealed that
personnel costs were the main driver of total college and university costs. Between 1961
and 2008, the Consumer Price Index increased 61 1.88 percent, while the higher education
price index increased by 967.19 percent. However, costs associated with personnel and
non-personnel categories increased at very different rates, with total Contracted Services
increasing 784.82 percent (moderately faster than the CPI during the same period, and
much slower than the aggregate HEPI increase of 967.19 percent), while total Personal
Higher Education and the Cost Disease 103
Compensation increased much more dramatically, for a total of 1022.57 percent between
1961 and 2008. This clearly demonstrates that personnel costs have been the main cost
driver of total costs in higher education since 1961.
An analysis of higher education costs and price increases to price increases
associated with other sectors of the economy also provides strong evidence to determine
that a cost disease impacts higher education. In order to determine whether a cost disease
impacts higher education, three of the most important comparisons involved assessing
higher education cost increases to the rate of price increases associated with PCE
Services and PCE Durable Goods, and more generally, PCE Services with PCE Durable
Goods. The theory of a cost disease suggests that cost and price increases associated with
service-related industries will be higher than those associated with the manufacturing
sector, since it is harder to leverage technology to increase productivity in service sector
compared to manufacturing-related industries.
Based on theory, researchers should expect that the prices associated with services
should rise faster than those associated with the manufacturing sector, represented in the
data set as Durable Goods. The analysis of the results strongly supports the presence of a
cost disease impacting the services in general, and higher education in particular.
Using the Tukey Post-Hoc Multiple Comparison Test, the mean difference
between price increases associated with PCE Durable Goods and PCE Services was
-198.51, which was associated with a significance level of ,036. This was below the .05
level of significance under which it was determined the null hypothesis should be
rejected, and suggests that price increases among the aggregated Services associated with
the NIPA PCE index are significantly higher than those associated with PCE Durable
Higher Education and the Cost Disease 104
Goods. This finding reveals that a general cost disease impacts the services relative to the
manufacturing sector, and provides an important proof of the first component of cost
disease theory.
However, in addition to assessing the extent to which a cost disease impacts the
services generally, this study also attempted to determine the extent to which a cost
disease impacts higher education in particular. This required comparing higher education
cost and price increases to those in associated with the Services.
As described in the literature review, many of the competing theories attempting
to explain higher education cost and price increases stress the unique role that the non-
profit status of the majority of colleges and universities has in propelling costs based on
competition for prestige and excellence. Indeed, most stress the impact of these factors as
driving costs associated with higher education's production function. If that is true,
researchers would expect to find that higher education costs and prices would increase at
rates significantly faster than those associated with other services, especially those
industries with profit-oriented firms.
However, Cost Disease theory suggests that cost and price increases associated
with the services should be similar, irrespective of the ownership structure associated
with a particular industry. Thus, to determine how well cost disease theory explains
higher education cost and price increases, higher education costs and prices were
compared against price increases in the Services. As described in the methodology
section, since higher education relies on multiple funding streams which can change over
time and which may impact tuition charges independent of changes in costs, it was
Higher Education and the Cost Disease 105
determined to compare both higher education costs and prices to price increases
associated with the service sector.
These findings also strongly support the idea that a cost disease impacts higher
education. Using the Tukey Post-Hoc Multiple Comparison Test, the mean difference
between price increases associated with PCE Services and Higher Education Cost
Increases was -82.395, which was associated with a ,816 level of significance. This was
well above the .05 level of significance under which it was determined that the null
hypothesis should be rejected, and suggests there is no statistical difference between cost
increases associated with higher education and price increases among the Service sector.
This finding reveals that, contrary to theories stressing idiosyncratic factors in fueling
higher intrinsic production costs, it seems higher education costs are more reflective of a
wider cost disease which impacts all services. This provides strong support for the theory
that the cost disease is primarily responsible for increasing higher education costs.
Theoretical Implications of the Research
The research associated with attempting to determine the extent to which a cost
disease impacts higher education costs and prices has important theoretical implications,
not only for public policy associated with higher education in particular, but also more
broadly with federal and state public finance. First, a cost disease associated with higher
education strongly suggests that higher education costs will continue to increase faster
than the rate of inflation as measured by the Consumer Price Index. Far from being the
anomalous result of the pursuit of excellence propelled by competition for prestige
among non-profit colleges and universities, the results suggest that higher education costs
are shaped by more general forces impacting the entire service sector. The inability to
Higher Education and the Cost Disease 106
leverage technology to increase productivity as fast as that associated with the
manufacturing sector will mean that higher education will face intrinsic cost increases
which outpace those associated with the general economy. This is not independent of the
ownership structure associated with college and universities, and related to higher
education's production function; its heavy reliance on labor as a key input in its
production process.
The cost disease in higher education has important implications associated with
the nature and extent of productivity increases which are possible among institutions of
education, and how these should he pursued. Moreover, since a cost disease implies that
college costs will outpace the rate of general inflation, the impact of a cost disease has
important implications for higher education finance. A cost disease affecting colleges and
universities implies that state and federal support for higher education should increase at
least as fast as the increases in the higher education price index on a per-capita student
basis. However, rapidly rising college costs threatens student access to higher education,
and impacts the mix of public and private resources used to finance a student's education,
including the relative mix of public versus private support, and even the current method
of providing loans to students and families. These policy implications will he explored in
more detail in the sections below.
In assessing the extent that a cost disease impacts higher education, the research
findings also determined that the Services are also impacted by a general cost disease.
This has important implications for public finance. A general cost disease suggests that
the cost of services will increase faster than those associated with the manufacturing
sector. Since the majority of local, state, and federal programs support the provision of
Higher Education and the Cost Disease 107
services, this strongly suggests that the cost of government programs will increase faster
than those associated with the general inflation rate.
This strongly suggests that the states and the federal government will face
continuing difficulty in raising revenue to support critical social programs even during
periods of relative prosperity, resulting in difficult resource allocation decisions in
budgeting among a variety of competing social claims. Determining the proper mix of
government services also implies that we face possible contentious social choices related
to the relative allocation of the economy between the public and private sector, whlch
may require increases in taxes to support the level of government services which are
deemed socially beneficial. The implications of a general cost disease on state and local
government finance will be also be explored in more detail in the sections below.
Increases in Higher Education Prices relative to Costs
One of the potentially puzzling research results revealed in Chapter N indicated
that while the rate of higher education cost increases was not statistically different from
increases associated with Services, the rate of increases associated with higher education
tuition sticker prices were significantly higher than those associated with Services. The
question remains why? The literature review identified two causes as most likely: burden
shifting and the growing reliance on tuition discounting as a way to target aid to highly
sought-after students. In fact, assessing the factors associated with undergraduate tuition
price increases is surprisingly difficult: colleges, even small ones, are multi-product firms
Higher Education and the Cost Disease 108
with multiple revenue streams, in which some "product lines" can be used to subsidize
others.
However, there does seem to be evidence that burden shifting has occurred in a
variety of important funding streams, beginning with the Federal Pell Grant. In fiscal year
1976, the maximum Pel1 Grant was set at $1,400, which on average could pay for 72
percent of the cost of attending a Public 4-year institution, and 35 percent of a private 4-
year college. By academic year 2006-07, the maximum Pell Grant was set at $4,050,
which by then could only pay for 32 percent of the average cost of a public 4-year
college, and only 13 percent of the average cost of tuition at a private 4 year college
(American Council on Higher Education, 2006). Moreover, the adjusted constant value of
the Pell Grant has not even kept pace with the general rate of inflation: the adjusted value
of the maximum Pell Grant awarded in AY 1976-77 in constant 2006 dollars would be
equivalent to $4,732, while the actual maximum Pel1 Grant available to students during
AY 2006-07 was $4,050 (American Council on Higher Education, 2006). The $4,732
figure is needed merely to keep up with inflation since 1976. If the Pell Grant was
adjusted to match true cost increases facing higher education based on the Higher
Education Price Index, the maximum Pel1 Grant for the Academic 2006-07 fiscal year
would have been $6,130 (calculation based on HEPI Index.) So, not only has the
maximum Pell Grant kept up with higher education costs associated with the HEPI, it has
failed to even kept pace with general inflation. This does suggest that, at least as it relates
to federal need-based student assistance, one of the causes for rapidly rising tuition prices
is burden shifting.
Higher Education and the Cost Disease 109
However, the bulk of public hnding for higher education comes from state
budget allocations. Here the data suggests there may be more complex forces at work:
Fairly significant shifts in funding depending on tax revenues resulting in fairly
significant fluctuations in the proportion of hnding derived kom tuition sources over
time.
Between 2002-03 and 2006-07, federal, state and local appropriations for public
degree granting institutions increased from $63.2 billion $73.9 billion in current dollars,
an increase of 17 percent (Table 352 Snyder & Dillow, 2010b); however, due to
enrollment growth, this translated to a per-student allocation of $ 6,840 in 2002-03 to
$7,780 in 2006-07, a per-student increase in current dollars of 13.7 percent (Table 352
Snyder & Dillow, 2010b). Meanwhile, between FY 2002 and 2006, the higher education
price index increased 17.1 percent, while the CPI increased by 11.6 percent during the
same period (based on HEPI Index). So while total public appropriations mirrored
increases associated with the Higher Education Price Index, public support failed to keep
pace with higher education costs on a per-capita student basis. This is a clear indication
that burden shifting occurred. Even during this relatively brief period, there were
fluctuations in the percentage of costs supported by tuition; tuition and fees accounted for
15.84 percent of operating revenues during the 2003-04 academic year, increasing to
16.97 percent of operating costs by AY 2005-06 (Table 352 Snyder & Dillow, 2010b).
The relative proportion of operating revenue provided by Tuition and fees actually
decreased to 16.67 during AY 2006-07 (Table 353 Snyder & Dillow, 2010~). Even with
these fluctuations, there does seem to be an overall trend on greater reliance on tuition
Higher Education and the Cost Disease 110
and fees over time; in 1980-81, tuition and fees accounted for 12.9 percent (Table 337
Snyder & Dillow, 2010a).
A similar pattern emerges when examining the proportion of revenue derived
from tuition for private non-profit institutions. When examining the period between AY
1997-98 through AY 2006-07, the proportion of revenue derived from tuition and fees
has declined, from an average of 27.82 percent from AY 1997-98 to an average of 26.03
percent in 2006-07 for all institutions. However, this masks large shifts based on the
investment returns associated with non-profit endowments. During AY 2001-02 and
2002-03, net endowment losses led to increased reliance on tuition and fees, from 38.1
percent of total non-profit revenue in AY 2001-02 to 39.72 percent of total non-profit
revenue in AY 2002-03, representing a large percentage increase in tuition and fees for
those institutions relying on endowments to offset tuition increases (Table 355 Snyder &
Dillow, 2010d).
A major problem associated with burden shifting is that it often occurs during
periods of economic strain and higher unemployment, resulting in declining state income
and sales tax revenue, or the declining value of portfolios associated with non-profit
institutions.
These shifts create an even greater burden on students and families precisely
when many students and families are hardest hit by economic forces beyond their control.
During these periods, when access to advanced education may serve as an effective and
relatively low-cost medium-term state and federal employment enhancement policy,
students are faced with an ever more imposing barrier to entry, involving
Higher Education and the Cost Disease 1 1 1
disproportionate increase in tuition relative to price increases associated with the rest of
the economy.
Tuition Discounting
The practice of tuition discounting also seems to be partially responsible for
escalating tuition sticker prices, particularly at private colleges. Using a sample of
institutions over time, the College Board estimated that the total undergraduate discount
rate among all sectors of higher education has increased between 1994-95 and 2004-05;
the discount rate for public 2-year institutions increased from 6.8 percent in 1994-95 to
12.5 percent in 2003-04 (a small sample size produced non-reportable results for the
2004-05 year); for public 4 year colleges, the discount rate increased from 11.7 percent
during the 1994-95 academic year to 14.7 percent by 2004-05. Meanwhile, discounting
was most prevalent at private 4 year colleges, with a discount rate of 23.8 percent during
the 1994-05 academic year, increasing to 33.5 percent by 2004-05 (Baum & Lapovsky,
2006). However, even these statistics may mask the growing reliance on tuition
discounting: the even faster rate of discounting associated with the entering freshman
cohort over time. According to an annual study conducted by the National Association of
Business Officers, the average discount rate for first-time, full-time freshman increased
from 27 percent in 1990 to 39 percent in 2002. Beginning in 2002, the average discount
rate remained around 38 percent. However, since then the discount rate climbed again
from 39.1 percent in the fall 2007 to 41.8 percent in fall 2008 (Fain, 2010). Meanwhile,
only 12 percent of the aid awarded as institutional grants was supported through
endowment income (National Association of College and University and Business
Officers, 2010). Additionally, more full-time entering freshman are also receiving these
Higher Education and the Cost Disease 112
awards, with the percentage increasing from 78.8 percent in 2000 to 82.3 percent in 2008
(National Association of College and University and Business Officers, 2010).
The growing reliance on tuition discounting seems to be part of a broader trend
toward awarding merit based aid through discounted dollars, which may have a much
more troubling side effect; limiting access to higher education. The NACUBO study
found that, among private non-profit institutions, only 36 percent of the of grants
awarded as institutional grant aid was based exclusively on financial need; 41.5 percent
of these awards were based on non-merit criteria, while 22.5 percent was based on a
combination of need-based and non-need criteria (National Association of College and
University and Business Officers, 2010).
In theory, tuition discounting can be used to facilitate need-based scholarships, by
charging wealthier students a higher tuition rate and shifting these resources to less well
off students. While some of the merit based grants are awarded to students with need, the
effect of the current practice of tuition discounting has been to siphon scarce institutional
resources away from need-based aid, awarding much of it to students and families with a
greater ability to pay for college education. Such practices, while potentially optimizing
from an institutional perspective collectively imposes extra barriers to needy students.
This practice is dt cross purposes with public policy goals of enhancing access and
college choice, and should be examined more closely. Assessing the extent to which
college prices, as distinct from college costs, are affected by tuition discounting and
burden shifting is also an important area for future research.
While a cost disease does seems to affect higher education costs, and the failure
of public sources of support to keep pace with higher education costs have also led to
Higher Education and the Cost Disease 1 13
rapidly rising tuition sticker prices, the widespread use and increasing reliance on tuition
discounting has also exacerbated the pace of tuition sticker prices. This also suggests that,
in addition to a cost disease, higher education prices may also be affected by some of the
factors identified by other scholars associated with non-profit institutions, including the
quest for excellence and prestige, and the competition this engenders. A more holistic
understanding of cost and price escalation among institutions of higher education may
combine elements of both these explanations. This will be explored more fully below.
Discussion and Analysis
There seems to be a paradox related to higher education, which has serious
implications for government policy: while the higher education sector is vital to
conducting cutting edge research, creating a well-educated workforce and enhancing the
productivity of the American economy, its internal production process seems to suffer
from the "productivity immunity" first described by Bruce Johnstone (1993), which
results in faster-than-inflation cost increases. These faster-than-inflation cost increases
have been observed since 1961, when the higher education price index was first
conceived, and have continued unabated though today.
The dilemma facing state and federal officials regarding higher education policy
is exacerbated by another finding of this study: that a general cost disease exists among
services relative to the manufacturing sector, in this study associated with durable goods.
Since most of the functions of state and federal government are associated with the
provision of services, this also means that costs associated with public sector will be
increasing faster than those associated with the general economy. This means that even in
Higher Education and the Cost Disease 114
relatively prosperous times, the states and the federal govemment face tough allocation
decisions in budgeting scarce resources among competing social claims.
However, this has become especially acute during the latest recession beginning
in 2008. In fiscal year 2010, the states faced budget gaps totaling $196 billion, in
response, 43 states have cut their state workforce, 30 states face cuts in their early
childhood and K-12 education program, 29 states face cuts in their public health budgets,
and 39 have planned cuts to support for their higher education sector ("States in the
Red,"). Meanwhile, during 2009, revenue to the federal government declined by 17
percent ($420 billion), while spending increased by 18 percent ($536 billion), creating a
budget deficit of $1.4 trillion. The resulting budget deficit equaled 9.9 percent of the
Gross Domestic Product, the largest deficit relative to the GDP since World War 11.
While decreasing slightly, the Congressional Budget Office estimates that the U.S.
Federal deficit will be $1.3 trillion for FY 2010 (Congressional Budget Office, 2010).
A general cost disease raises severe challenges not only for higher education
finance in particular, but more generally for the financing of govemment programs and
those goods which are seen as socially beneficial: since the bulk of government programs
are service-based, a general cost disease associated with services suggests that the cost of
government will continue to rise faster than general inflation. Significantly, along with
higher education, health care and K-12 education have been identified as services also
facing a cost disease, compounding the relentless pressure on government budgets
(Baumol, 1967, 1993; Gundlach et al., 2001; Snower, 1993).
Nor is the United States alone in facing these challenges in funding public
services; the lingering effects of the 2008 recession has placed severe stains on the
Higher Education and the Cost Disease 115
budgets of many European Union member states (Erslanger, 2010), particularly support
for higher education (Labi, 2010). The escalating costs associated with the provision of
services have also been linked to a general cost disease facing other OECD countries as
well (Gundlach et al., 2001; Snower, 1993).
While the services are not devoid of any productivity grains, these will be lower
than those associated with the manufacturing sector. The Bureau of Labor Statistics
estimated that productivity in the manufacturing sector increased at an annual average
rate of 4.0 percent between 1990-2000, and by an annual average rate of 3.7 percent
between 2000-2007; in contrast, productivity increases in the n o n - f m business sector
which includes a wide number of businesses, increased at an average annual rate by 2.1
percent between 1990-2000, and 2.6 percent between 2000 and 2007 (Bureau of Labor
Statistics, Labor Productivity and Costs).
As long as the general economy experiences productivity increases, society
should be able to support its government programs. However, this will involve potentially
vexing and socially contentious issues in determining the relative portions of the
economy which should be allocated between the private and public sector, and may
require increased taxation to support the level of government services which are deemed
socially beneficial. This includes support for higher education.
In this environment of severely constrained resources, the higher education
community cannot simply expect revenue enhancements based on prior good will or past
services rendered. Public support for higher education, both through direct subsidies to
public institutions, and indirect support provided as need and merit-based aid to students
must be centered on the benefits which higher education provides to the rest of society.
Higher Education and the Cost Disease 1 16
Public Funding: Finding the Balance between Private Benefits and Public Positive Externalities:
The general argument for providing public support for higher education is based
on the classic case of public externalities: The education of a highly skilled work force
provides benefits to society beyond the personal benefits received by the individual
student, although these are considerable, typically including increased earning potential, a
reduced likelihood of unemployment, and, if unemployment occurs, less time spent
among the unemployed.
The greater the extent to which society benefits through the creation of highly
skilled workers who enhance the competitiveness and productivity of the workforce, the
stronger is the argument for public support. In this case, a higher education sector left
strictly to market forces would produce a socially inefficient amount of students receiving
advanced degrees, since this market would be based strictly on the private benefits
received by particular students. Public investments in higher education are needed to
ensure that a socially optimal amount of higher education is produced, based on the
residual benefits received by society independent of private benefits received by students.
Blundell, Dearden, Goodman, and Reed (2000) have further categorized these
benefits into three distinct groups: (a) private financial returns, defined as the extent to
which an advanced degree improves earning potential or jobs prospects, (b) private non-
financial returns, which encompasses the benefits an individual receives not measured by
earnings, such as more desirable and interesting work, and (c) Social returns, which
Higher Education and the Cost Disease 117
defines the benefits that Ingher education provides to other members of society beyond
the private retums received by degree recipients.
The social retums to education are substantial, including enhanced productivity, a
highly skilled workforce, a larger tax base, and broader civic responsibility (Camegie
Commission on Higher Education, 1973; Carol1 & Emre, 2009). Moreover, recent
projections on the future needs of the US. workforce indicate that advanced education
will be even more critical to U.S. economic prosperity. A recent research study conducted
by the Georgetown University's Center on Education and the Workforce indicated that 63
percent of all jobs in the US economy will require at least some college-level education
by 2018, up from just 28 percent in 1973 (Camevale, Smith, & Strohl, 2010). Based on
the structural changes transforming the economy, Camevale, Smith and Strohl(2010)
estimate the US . will need an additional 22 million workers with a college degree;
however, based on current college attainment rates, they project a gap of 3 million degree
recipients required by the economy. This will pose serious challenges to American
prosperity and the ability to compete in a rapidly globalized knowledge economy.
The social benefits derived from higher education have important policy
implications on the agpropriate level of public investment which should be allocated to
higher education, as well as on the appropriate mechanism through which student
education should be funded. These issues will be explored along with other policy
recommendations in the next section.
Higher Education and the Cost Disease 1 18
Policy Recommendations
Funding for Higher Education Nced-Based Aid Programs
The evidence found in this study reveal a number of issues which hopefully can
help guide public policy associated with higher education. First, the rate of increases
associated with lngher education costs is not statistically different from the rate of price
increases associated with Services included in the National Income and Product
Accounts. Furthermore, the aggregate rate of increases associated with all NIPA Services
was significantly statistically higher than the rate of increases associated with NIPA
Durable Goods. Significantly, most of the services tracked as part of the NIPA Index
were supplied by profit-oriented industries; this suggests that the all services, irrespective
of the ownership structure of the firm are affected by a cost disease, higher education
among them. The results described in Chapter IV provide strong evidence that the rate of
cost increases associated with higher education have not been excessive, nor based on the
factors idiosyncratic to non-profit institutions of higher education.
This strongly supports the argument that both states and the federal government
should attempt to increase public investment in higher education, especially support
associated with need-based aid programs, at the rates which approximate increases in the
higher education price index, on a per-capita student basis. The extent of government
support for higher education will be based on the social benefits derived from higher
education. As indicated above, these are substantial, and are spread across the degree
earner's lifetime and involve sigmficant benefits to all taxpayers. Social benefits derived
Higher Education and the Cost Disease 119
from enhanced degree attainment include larger tax payments based on higher wage
earnings, less need for social support programs and consequent government transfer
payments, as well as reduced likelihood that an individual will engage in criminal activity
(Carol1 & Emre, 2009).
The Obama administration has highlighted the importance of advanced education
in achieving broad national objectives, and has recommended a considerable
enhancement in the Federal government's investment in its support programs for needy
students: the President's 201 1 Federal Budget calls for a 29.2% increase in the Pell Grant
program, which provides a national floor for need-based aid ("Highlights of Obama's
Fiscal 201 1 Budget for Higher Education and Science," 2010). Moreover, the President
has proposed making the Pell Grant an entitlement, which would be allocated an
automatic increase each year based on the number of students who qualify. Finally, in a
signal which may reflect an intrinsic understanding of the cost disease facing higher
education, the President has proposed increasing the maximum Pell Grant each year by
one percentage point above the rate of inflation (Basken, 2010).
However, both the States and the Federal government face an extremely crowded
legislative agenda, with many competing claims among extremely beneficial social
assistance programs. Institutions of higher education must earn the public's trust and as
well as their financial support by working to minimize institutional policies which act at
cross purposes with state and federal policy objectives. This includes the heavy reliance
on tuition discounting.
Higher Education and the Cost Disease 120
Institutional Tuition Discounting and Public Policy
As described above, the use of tuition discounting is both widespread and
increasing. In 2009, the average aggregate discount for the entering freshman class of
full-time students was over 40 percent (National Association of College and University
and Business Officers, 2010). While mostly observed at private institutions, the use of
tuition discounting has been increasing among public institutions as well (Baum &
Lapovsky, 2006).
Some may question why this practice would impact public policy, since
institutions are leveraging their tuition revenue to help shape the academic character of
their entering college classes. However, a strong guiding principle of state and federal
higher education policy is to maximize the scarce dollars invested in higher education to
increase access to college, particularly for disadvantaged students, along with
maximizing student choice in deciding on attendance.
There is a legitimate concern that tuition discounting, which in practice often
awards large merit scholarships to students with desired academic credentials, and who
are often from more well-off families, dissipates the effectiveness of federal and state
need-based awards. Needier students receiving state or federal aid may in fact face a
higher institution-discounted tuition price (before federal and state aid are accounted for)
than less needy students. Seen in this way, state and federal need-based aid awards have
the unintended consequence of subsidizing the institutional merit-based policies, reducing
the effectiveness of state and federal efforts. This practice has been criticized as one
which in effect helps reduce the enrollment of low-income students in colleges and
universities (Kean, 2006). While it is outside the scope of this study to examine the extent
Higher Education and the Cost Disease 121
to which federal and state need-based aid awards are "captured" by institutions, the
practice erodes public confidence in the institutional commitment to broader social
purposes. The Obama administration has recognized the importance of increased
enrollment in achieving national objectives, the President (as cited in Nelson, 2010)
offered a challenge to institutions of higher education during his 2010 State of the Union
address, stating: "In the United States of America, no one should go broke because they
chose to go to college. And by the way, its time for colleges and universities to get
serious about cutting their own costs - because they have a responsibility to help solve
this problem" (Nelson, 2010). While the public commitment to higher education has been
generous, this should not be viewed as a bottomless pool of resources; public patience is
wearing thin, and practices such as tuition discounting erode public support.
Increasing Productivity
The presence of a cost disease implies that the rate of cost increases associated
with services will be higher than those associated with the manufacturing sector.
However, this does not mean that productivity increases are impossible, nor is the rate of
productivity gains constant over time, even in the service sector. New technology or
different approaches to output can lead to increases in productivity. Increasing
productivity in higher education is an important concern worth exploring. However, this
must mean more than merely placing more students in class sections, or even changing
the allocation mix of faculty time among teaching education and research, as is often
suggested.
While increasing the ratio of students to faculty would provide immediate
"productivity gains", this would surely lead to a decrease in quality over time, reflecting
Higher Education and the Cost Disease 122
the classic dilemma associated with a service sector where labor itself is the end product
of output. However, assessing the appropriate balance between the ratio of faculty to
class sizes for various academic programs of study and courses within programs may
provide a baseline to help colleges and universities manage the allocation of staff. This
can provide some useful insight as a basis for discussion on institutional staffing, but is
not an end in itself to staffing strategy.
Nor is the reallocation of faculty time among the components of teaching,
research, and service in itself a valid method of increasing faculty "productivity". As
Johnstone (1993) has argued, increasing the amount of time faculty at research
universities engage in teaching at the expense of research will not make them more
productive, but merely reassign them to a different job involving relatively less research.
Moreover, to the extent that the research conducted at colleges and universities enhances
national prosperity and economic competitiveness, such a reallocation would likely harm
overall economic output over the longer term.
However, assessing the optimal balance of teaching, education, and research may
involve different time allocations for particular faculty members over time. Academic
departments should actively assess the research and service productivity of its faculty,
which may lead to relative shifts in the time allocations toward greater teaching as faculty
research output changes. While this would lead to increased efficiency of the use of
faculty time, this is unlikely to increase overall faculty productivity, merely a reallocation
of time to different productive outputs (Johnstone, 1993).
Increasing Productivity: Technology and On-Line Classes, Possibilities and Limitations
Higher Education and the Cost Disease 123
Many people have recognized the link between increasing productivity and
reduced costs, and have suggested ways to increase productivity in producing higher
education as a way to stave off cost increases (Vedder, 2007). This has proved to be
somewhat elusive, although many have focused on the possibility of leveraging new
technology as a way to reduce the intrinsic costs associated with providing higher
education services.
However, it seems there are two distinct sources of cost savings associated with
the use of on-line education, which have become muddled together, and a possible source
of confusion. Much of the use of this technology as a method of reducing the cost of
delivering education to students is not associated with the delivery system, but with mix
of personnel contracted to lead online classes, and the sectors of the educational
community which have most embraced the new technology.
During the last several decades, a major shift has occurred in the employment
practices at colleges and universities, partly in response to rapidly accelerating costs; the
increase in the number of contingent and part-time faculty relative to number of full-time
and tenure-track positions (Schuster & Finkelstein, 2006). In 1970, 474,000 instructional
faculty were employed in degree granting institutions, including 369,000 full-time and
104,000 part-time instructional staff, reflecting a full-time percentage of 77.9 percent. By
2007, a total of 1,371,000 faculty were employed by degree-granting institutions,
however, only 703,000 (51.3 percent) were employed full-time, while 688,000 were
employed on a part-time basis @ES, 2009, Table 249). While statistics on the type of
faculty teaching online classes are difficult to find, it seems that many of the faculty
associated with teaching online classes are part-time; this is particularly true of the sector
Higher Education and the Cost Disease 124
of higher education which has most embraced the new technology, for-profit colleges and
universities (Morey, 2004). This is certainly true of the largest for-profit university, the
University of Phoenix. The University of Phoenix does not employ any tenured faculty.
In 2004, it employed only 285 full-time faculty, and 17,000 part-time faculty. Four
thousand of the part-time faculty were employed in its University of Phoenix online
program (Morey, 2004).
Moreover, even among full-time faculty, there is a large difference in pay
between non-profit private and public institutions and profit-oriented higher education
institutions. During the 2008-09 academic year, the average faculty salary of U.S. full-
time faculty was $73,570; this included an average salary for employees at public non-
profit institutions of $71,237, $79,358 for those employed at private non-profit colleges
and universities, and $52,557 for full-time employees at for-profit institutions (DES,
2009, Table 259).
Thus, when assessing the potential of technology to increase productivity and
reduce costs in delivering education to students, it is important to first isolate the
potential savings to institutions of higher education based on using part-time and
contingent faculty as opposed to cost savings exclusively associated with the enhanced
capabilities of online technology.
A similar problem related to productivity gains in the use of computer based
education involves class sizes of online classes. The researcher was unable to locate
statistics identifying the size of online classes, however, while content delivery
technology may facilitate increasing enrollment in online classes, these "productivity
gains" would be similar to enrolling more students in traditional classrooms as a way to
Higher Education and the Cost Disease 125
decrease unit costs. Future research assessing the relative class-size of on-line and in-
person course sections would provide more up-to-date estimates of how this new
technology is being used, and provide aid in the assessment of academic quality of online
versus face-to-face teaching.
However, the potential for genuine cost savings using computer based content
delivery are real, although probably smaller than suggested, once class size and the
employment status of professors leading on-line classes are accounted for. The most
immediate savings involves infrastructure costs: using computer technology, course
delivery can be separated from the extremely expensive infrastructure and maintenance
costs associated with traditional brick-and-mortar college and university classroom space.
While the delivery of online courses involves initial added investments in the
technological infrastructure necessary to enable interactive exchanges among students
and professors, these are relatively inexpensive compared to huge investment in facilities
required to teach face-to-face classes.
This has become a critical concern as state budgets have failed to keep pace with
per-capita increases in FTE student enrollments, and more recently, deficits have forced
cutbacks to higher education across a large number of states, as described above (Hebel,
2010). The infrastructure costs to support traditional higher education are substantial; in
current dollars, combined capital appropriations and capital grants and gifts associated
with public degree granting institutions increased from $7.96 billion during the 2002-04
academic year to $ 10.84 billion during AY 2006-07 (DES, 2009, Table 352.) While
online courses will not eliminate the need for investment in facilities and other
infrastructure, it may reduce the rate of cost increases since, even for traditional brick-
Higher Education and the Cost Disease 126
and-mortar institutions, not all students would be required to be seated in campus
facilities to the same extent.
Increasing ESficiency
The cost of completing undergraduate associate and bachelor degrees are high,
both from the standpoint of student and families and the resource investment by society.
Increasing the efficiency associated with degree attainment provides one of the most
important ways to increase the overall productivity and reduce costs associated with
colleges and universities. Johnstone has described this as increasing the ''learning
productivity" of American higher education (Johnstone, 1993).
Four year degree completion rates have somewhat increased for first-time full-
time undergraduates attending 4-year colleges, from 33.7 percent for the entering 1996
cohort, to 36.2 percent for the Fall 2001 cohort. The 6 year graduation rate has also
slightly increased, fiom 55.7 percent for the fall 1996 cohort to the 57.3 percent for the
fall 2001 cohort (see Table 10).
Table 10.
Percent of AN First-time Full-time Entering Students at 4-Year Institutions Completing Undergraduate Degrees: Fall I996 Through Fall 2001 Cohorts
Within 4 Within 6 Entering Cohort Years Within 5 Years Years 1996 Cohort 33.7 50.2 55.4 1997 Cohort 34.1 51.1 56.0 1998 Cohort 34.5 51.5 56.4 1999 Cohort 35.3 52.3 57.1 2000 Cohort 36.1 52.6 57.5 2001 Cohort 36.2 52.6 57.3 Source: Digest of Educational Statistics 2009, Table 331
Higher Education and the Cost Disease 127
Meanwhile, degree completion rates for students attending 2 year colleges have
slightly declined; the 150 percent certificate or associate degree completion rate for full-
time, degree seeking students for all institutions decreased from 29.3 percent for the
1999 Starting Cohort to 27.8 percent for the 2004 starting cohort (see Table 11).
Table 11.
150 Percent Certiififate or Associate Degree Completion Rate for Full-time Degree Seeking Students: 1994 Through 2004 Starting Cohorts
All 2 Year Public Private Non- Entering Cohort Inst Institutions Profit 1999 Starting Cohort 29.3 22.9 44.7 2000 Starting Cohort 30.5 23.6 50.1 2001 Starting Cohort 30.0 22.9 54.8 2002 Starting Cohort 29.3 21.9 49.1 2003 Starting Cohort 29.1 21.5 49.0 2004 Starting Cohort 27.8 20.3 44.4
While increasing, the 6-year degree completion rates for first-time full-time
students are still less than 60 percent, an attrition rate which represents considerable
expense for those not completing a degree, and a huge investment of public and private
resources. Rapidly rising costs may also compound the need for lengthened study, since
some students may be required to reduce academic their course load or even temporarily
drop out to earn money to finance their college education (Johnstone, 1993).
Online educational technology may provide critical assistance in shortening
degree attainment for students. The technology, combined with a larg'e number of
institutions delivering course content, can increase student access to courses, allowing
them to complete required courses with greater ease and convenience (Allen & Seaman,
Higher Education and the Cost Disease 128
Increase High School Proficiency
It is much more expensive to teach students in postsecondary institutions than in
high school. However, an increasingly large number of college students are required to
take remedial coursework, particularly in math and English, due to deficiencies in high
school preparation, with additional costs. As Terry Hartle, the Senior Vice President for
Government and Public Affairs at the American Council on Education explained, "If
you're academically prepared for college, you're far more likely to graduate. Remedial
education is expensive and inefficient, and if we're able to reduce it, we'll be able to
focus on college-level work" (Sewall, 2010).
This is a complex issue involving the educational pipeline from high school to
college, perhaps even earlier. Since access to higher education is a primary mechanism of
equality or opportunity, colleges and universities must no doubt continue to afford the
opportunity of remediating deficiencies among entering students. However, this issue
must be assessed holistically in order to increase the effectiveness of resources allocated
at all levels of the educational pipeline. In an era of constrained resources, all levels of
education must do a better job at education to ensure students have the appropriate level
of knowledge as they advance between different levels of the education system. Having
students achieve the appropriate level of pre-collegiate education before entering college
would provide significant system-wide cost savings, and may even enhance high school
graduation rates as well.
The newly released "Common Core State Standards," developed by the National
Governor's Association Center for Best Practices and the Council of Chief State School
Offices in June, 2010, may provide an important framework to achieve these objectives.
Higher Education and the Cost Disease 129
One of the stated goals of the new standards are to incorporate college and career
standards into the K-12 curriculum (Common Core State Standards Initiative, 2010). In
addition to enhancing the education of students in the K-12 system, the new standards
may provide an important mechanism in creating a rational educational transfer
throughout an extended K-20 educational pipeline, increasing total system-wide
efficiency and reducing the costs of remediation within college.
College Credits in High School
Students are able to earn advanced placement credit in high school which may be
able to be applied in college; however, the opportunity to earn these credits is often
limited to the most advanced high school students. Proposals for expanding the
opportunity for a greater proportion of high school students to earn college credit are not
new (Johnstone, 1993). However, this has taken on greater significance with rapidly
rising college costs, and new international models on how these partnerships can work.
The ability for high school students to earn college credit in high school would create
greater efficiency in the utilization of resources in colleges and universities. As Johnstone
indicated in 1993, increasing learning productivity requires greater collaboration between
high school and institutions of higher education. With rapidly rising college costs,
fostering additional opportunities for high school students to earn college credits take on
added significance. By improving the articulation of coursework between high school and
college, the proposals associated with the K-12 Common Core Standards Initiative may
not only improve college-level readiness in high schools, but may also be used to allow a
greater proportion of high school students to earn college credits.
Higher Education and the Cost Disease 130
Three Year Undergraduate Degree
The idea of earning an undergraduate degree in 3 years is not new; Judson
College, located in Alabama has offered a 3-year degree for over four decades
(Alexander, 2009), and Johnstone (1993) suggested redesigning the academic calendar to
facilitate year-round learning. However, the rapidly increasing costs associated with
higher education and the failure of government support to not only keep pace with
increases in the Higher Education Price Index, but also recent cutbacks in state support
have added to the impetus to increase institutional productivity by utilizing facilities more
effectively, and reducing the time required to compete an undergraduate degree.
In a New York Times editorial (Trachtenberg & Kauvar, 2010), Stephen Joel
Trachtenberg, the President emeritus of George Washington University, supported the
idea, suggesting that all colleges should strongly consider creation of a 3-year degree
program, based on the per-student cost reduction, and the possibility to maximize the use
of institutional infrastructure:
Three-year curriculums, which might involve two-full summer of studies with short breaks between terms, would increase the number of students who could be accommodated during a four-year period, and reduce institutional costs per student. While there would be costs for the additional teachers and staff, those would be offset by an increase in tuition. Meanwhile, institutions that go quiet in the summer, incurring the unnecessary expense of running nearly empty buildings, would be able to use their facilities year-round. (Trachtenberg & Kauvar, 2010)
In addition to the increased degree efficiency and productivity associated with a 3 year
degree, there may be additional benefits: the ability to more effectively compete with
higher education institutions in the European Union and those included in the European
Higher Education Area, which have standardized a common European degree across
counties member states, which also utilize a 3 year degree (Adelman, 2008,2009).
Higher Education and the Cost Disease 131
This process, which began in 1999, now includes the institutions of higher
education across 46 nations, also including the United Kingdom, Russia, and Turkey.
Meanwhile, Australia, New Zealand, China, and India have also been closely monitoring
the EHEA framework, and 18 nations in Latin America are developing the "Tuning
Model," the portion of the Bologna process which attempts to align goals among
academic disciplines (Labi, 2009). A unified degree structure based on a standardized
three-year undergraduate degree, but which also links degree knowledge and attainment
to employment, will pose an increasingly appealing competitor to the traditional model of
higher education historically offered in the United States. A 3-year degree many not only
enhance productivity and learning efficiency, but also help US. institutions of higher
education more effectively compete in a rapidly globalizing educational marketplace.
Financing of Higher Education
The principle finding of this study, that higher education costs are affected by a
systematic cost disease, has important implications for the financing of higher education
for both states and the federal government. Much of the current policy debate associated
with financing higher education has been focused on the reason for rapidly rising costs
and tuition sticker prices. As indicated in Chapter IV, cost increases associated with
higher education are not statistically different from price increases associated with
Services tracked in the National Income and Product Accounts. Most of the NIPA
Services are associated with profit-oriented industries, which suggest that much of the
increasing higher education costs are not as easily controlled as many critics contend.
This provides a strong argument that federal and state support for higher education
should increase at the same rate as the Higher Education Price Index, on a per-capita full-
Higher Education and the Cost Disease 132
time equivalent student basis, to the extent possible. Obviously, both the states and the
federal government face many competing social claims on the allocation of public
resources, which have to be weighed carefully in distributing public funds. This also
means that public resources whch are allocated to higher education should be targeted to
maximize their utility and effectiveness. This leads to additional considerations, which
are explored below.
Allocation of Higher Education Public Funding: Institutional Subsidies, and State Need and Merit-based Student Assistance Grants
Since public resources are scarce, all federal and state funds should be allocated to
achieve maximum effectiveness. This raises the question of how effectively public
funding is currently being utilized to achieve state and national policy objectives.
State Need versus Merit Grants
Almost parallel to the growing use of tuition discounting in the awarding of
institutional merit based grants, there has been an increasing tendency on the part of the
states to offer merit-based as opposed to need-based grants. The percentage of full-time
dependent students receiving state grants increased from 14 percent during AY 1992-93
to 28 percent during the 2008-09 academic year (Baum, Payea, & Steele, 2009). The
average grant per recipient increased from $2,350 in AY 1992-93 to $3,130 during AY
2007-08 in constant 2007 dollars. However, the proportion of aid distributed using need-
based criteria has declined significantly, from 90 percent in 1992-93 to 72 percent in
2007-08. Meanwhile, during the 200-08 academic year, 47 percent of students coming
from families with parental income less than $ 32,500 received an average grant of
Higher Education and the Cost Disease 133
$3,400, while 13 percent of students coming from families with parental income of
$100,000 or more received an average state grant of $3,000 (Baum el al., 2009).
The primary principle in awarding government assistance grants to college
students is to broaden access to a college education to create equality of opportunity
(Camegie Commission on Higher Education, 1973). While there is no doubt justification
for the awarding merit-based state grants, enhancing access to education is best achieved
through need-based grants to the neediest students.
This is particularly important in the context of rapidly rising costs, where state
support has failed to increase at the rate associated with the higher education price index
on a per-capita student basis. While many of the state merit-based grants are no doubt
awarded to students who have some need, these outcomes are incidental to the method of
awarding merit-based grants. Much of this non-need grant aid, particularly to students
from families in the highest income brackets, is awarded to students who would have
attended college regardless of this financial support. In an era of constrained public
resources devoted to higher education, this dissipates the effectiveness of state financial
assistance. The most effective use of direct student assistance grants is in the form of
need-based aid, with the purpose of increasing college participation among students from
family income brackets with historically lower college attendance. By reducing the
opportunity cost of attending college for needy students, these targeted need-based grants
would be more effective in inducing students who might otherwise not attend college to
enroll, increasing the aggregate social benefits of college attendance.
Higher Education and the Cost Disease 134
Institutional Subsidies vs. Student Need-based Grants
As indicated above, the primary justification for a public investment in higher
education stems largely from the social benefits achieved through broader participation in
undergraduate education. Currently, there are two primary mechanisms for public support
for higher education: direct institutional subsidies as well as grants (both need and merit-
based) and loans to students.
The vast majority of state assistance involves direct subsidies to public
institutions. This translates into significantly cheaper tuition charges at public institutions.
In 2008-09, the average tuition and fees for all public institutions was $12,113, while the
average tuition and fees for non-profit institutions was $31,921 (DES, Table 334).
Special support for public institutions is desirable to the extent that public
institutions serve particular state interests associated with community service and
extension activities, continuing education, and research (Carnegie Commission on Higher
Education, 1973).
However, many of the social benefits achieved through college attainment accrue
to society irrespective of whether a student attends a public or private institution
(Carnegie Commission on Higher Education, 1973). Moreover, the general subsidy
provided to students attending a public institution accrues both to students from wealthy
as well as needy families. An argument could be made that this dissipates the
effectiveness of direct institutional subsidies in increasing attainment rates, since some of
the funding allocated to public institutions accrues to students with relatively greater
ability to pay.
Higher Education and the Cost Disease 135
Given rapidly rising costs affected by a cost disease, and constrained public
resources allocated to the higher education sector, it seems a relative reallocation of state
aid away from direct subsidies to institutions and toward increased need-based aid would
better maximize the limited h d s committed by the states to increase access and
opportunity to those students experiencing relatively greater need.
Income Contingent Loans
With rapidly rising college costs and uneven state and federal grant support, the
aggregate amount of student loan debt has been increasing substantially. To gain
perspective, in constant 2008 dollars, the amount of money awarded through the Pel1
Grant program increased from nearly $9.74 billion in academic year 1998-99, to $18.2
billion in academic year 2008-08, an increase of 87 percent. State grants (both need and
merit-based), increased from $4.95 billion during the 1998-99 academic year, to $8.492
billion during the 2008-09 year (in constant dollars), an increase of 72 percent (Baum et
al., 2009).
During the same period, subsidized loans (loans for which the federal government
pays the interest while a student maintains at least half-time residency in college)
increased by only 45 percent (from $21.981 billion in AY 1998-99, to $31.95 billion in
AY 2008-09. Meanwhile, unsubsidized loans (for which the student either pays the
interest rate during college or it is capitalized into the loan's principle) increased by 165
percent (from $14.691 billion during AY 1998-99 to $38.9 billion in AY 2008-09 (Baum
et al., 2009).
However, even this masks the great shift in loan-based support; during the same
period, PLUS loans, federal loans offered to parents based on their credit-worthiness,
Higher Education and the Cost Disease 136
increased 194 percent, from $3.985 billion in AY 1998-99 to $1 1.732 in AY 2008-09,
while non-federal altemative private loans, often charging high interest rates, increased
from $3.91 billion in AY 1998-99 to 11.9 billion by AY 2008-09, an increase of 204
percent (Baum et al., 2009).
Two-thirds of students graduating with a college degree incurred some type of
personal loan debt; student debt has been increasing approximately six percent per year
since 2003-04, with an average student loan of $23,200 for the class of 2008 (Reed &
Cheng, 2009). However, this debt picture is highly deceptive, since it accounts for only
student-based loans, and fails to account for the total family debt which includes rapidly
rising parent PLUS loans and non-federal alternative loans, the fasting growing portions
of all college-related loans. Other than aggregate loan volumes described above, the
researcher has been unable to uncover loan debt averages which account for all sources of
loans.
This alarming rise in student debt may act as a deterrent for qualified students to
attend college, particularly for students from low income families. With rapidly rising
costs affected by a cost disease, and loans a growing component of student aid, it is vital
to place the system of student loans on a more rational footing: this involves moving
away from the current method of mortgage-style student loans toward more student-
friendly income-contingent system of student loans.
There is strong theoretical support for some system of student loans; since the
student is a primary beneficiary of the education s h e receives, students should pay for at
least some part of their education. Loans are available for students and their families who
lack the capability to finance a student's education from current resources. However, it is
Higher Education and the Cost Disease 137
extremely difficult to create capital markets surrounding the provision of human capital,
since it is impossible to collateralize a person's education. Due to these risks, capital
markets may not be willing to provide a sufficient level of financing for education at
socially desirable levels (McPherson & Shapiro, 1991).
Moreover, there is a problem associated with risk: while in the aggregate, a
college education is a very good investment, the returns fluctuate wildly based on a
number of personal idiosyncratic factors, in addition to the type of degree a student earns.
Additionally, students failing to complete a degree would still face steep loan
repayments. In these circumstances, a student financing an education primarily through
loans would be hobbled by substantial debt, which could prevent himiher from achieving
other aspirations such as purchasing a home. Facing these risks, a number of college
capable students may choose not to attend college, even though they could benefit
substantially from a college education (Chapman, 1997; McPherson & Shapiro, 1991).
Income contingent loans may provide a mechanism which ensures students pay
for some of the personal benefits which accrue to them from earning a college education,
but which minimize these barriers to student participation. This concept is not new, the
idea was first proposed by Milton Friedman in 1955 (Congressional Budget Office,
1994). While they can be implemented in a variety of ways, the basic component of
income contingent loans is that repayment is based on income earned over a period of
time, so the amount repaid is variable based on the income of the borrower. Depending
on the plan, the amount borrowed is paid back over a specified time period, , usually
from 12 to 20 years. So, unlike mortgage-type loans, which involve a fixed repayment
schedule regardless of the borrower's income, the amount paid back depends on the
Higher Education and the Cost Disease 138
future earnings of a student. This greatly reduces the uncertainty and risk associated with
borrowing, since payment is tied to one's ability to repay, not a unalterable payment
schedule (Barr, 1993).
Australia is the country which has most embraced this repayment system. Facing
both an improving high school retention rate, along with a demographic increase in the
number of college bound students, Australia's higher education system went through a
rapid expansion. The Australian government could no longer afford to maintain the free
tuition system first created in 1974. However, the government did not want the new
financing system to become a barrier which would discourage enrollment, especially for
the economically disadvantaged (Chapman, 1997).
Beginning in 1989, Australia introduced the Higher Education Contribution
Scheme (HECS). Under the new system, students have the option of paying the HECS
fee when initially enrolling at an institution, at a discounted rate of 25 percent, or the
HECS fee is deferred until after graduation. Payments are then automatically collected
through the tax system as they begin earning income after graduation. However, students
were charged the full cost of their education; under the initial framework, the student
income contingent loan was equal to 20 to 25 percent of the actual subsidy provided to
students (Chapman, 1997).
The system has a number of important benefits worth considering. First, it
removes the tremendous burden associated with the repayment of student loans after
graduation. Second, it provides default insurance for student borrowers, removing the
uncertainty associated with borrowing for a college education.
Higher Education and the Cost Disease 139
The Obama Administration has taken the greatest step toward introducing income
contingent loans for students, although still in very limited form. Beginning in the 2009-
10 academic year, students were offered an income contingent loan repayment option for
Federal Direct subsidized and unsubsidized loans, excluding PLUS (Parent) loans. Under
the plan, repayment is based on adjusted gross income, family size, and total loan
amount. There are two methods of calculating repayment: a repayment rate based on a 12
year schedule, multiplied by an income percentage factor which varies by income, or 20
percent of discretionary income. If payments have not covered the interest which has
accrued on the loan, unpaid amounts are capitalized once a year. The maximum
repayment period is 25 years, after which the loan is discharged (Department of
Education, 2010).
While a start, the U.S. version leaves much to be desired; it is complex and
confusing, involving the potential capitalization of interest, and still relatively high
annual repayment amounts compared to the Australian plan. Perhaps worst of all, it only
covers direct loans, the portion of student and parent loan debt which has been rising
relatively slowly. It does not include Parent or alternative loans as an option for income
contingent loans.
While the Obama administration has standardized portions of the loan system,
eliminating private lenders associated with its subsidized and unsubsidized loan system,
the US. student loan system has become a serious obstacle to expanding access to higher
education. As described above, the terms of repayment are still complex and confusing,
intimidating even to parents with prior experience. Students and parents are facing larger
loan debts, with PLUS and alternative loans increasing at an alarming rate. Limiting
Higher Education and the Cost Disease 140
student risk with a rationalized system of income contingent loans will be an important
method of increasing access to higher education, vital if the US is to compete in an
increasingly globalized knowledge economy.
Administrative Salaries and the Costs of Regulation
An examination of the components of the higher education price index revealed
that, while both faculty and administrative salaries increased faster than the aggregate
HEPI index, the costs associated with administrative salaries rose even faster than faculty
salaries. Adding even greater cost pressures on institutions, the number of administrative
positions increased far faster than faculty positions. The proportion of managerial and
executive positions increased &om 5 percent in 1976 to 6 percent in 2007, while the
proportion of non-teaching professional staff increased from 10 percent in 1976 to 20
percent in 2007 (DES, 2009, Table 244).
It is outside the scope of this study to determine the causes for the rapid rise in
administrative positions. Many of the positions, no doubt, reflect the added technological
complexity surrounding the delivery of education, such as technology managers and
educational specialists. However, it seems many of the positions may be associated with
greater accountability and oversight increasingly required by both the federal and state
governments, as well as accrediting agencies. One former university president estimated
that regulatory requirements cost his university 7 percent of all tuition revenue
(Alexander, 2009). This suggests there may be a tension associated with the current
regulatory climate, increasingly focused on assessment of outcomes and accountability of
resources; the added scrutiny to account for the public resources invested in colleges and
Higher Education and the Cost Disease 141
universities, and the additional requirements associated with process and outcomes
assessments, may actually be a cause for additional expenditures.
The rapid growth in administrative and professional non-teaching positions and
the potential association to additional administrative oversight requires additional
research; this also suggests that government regulatory bodies and accrediting agencies
must strike a careful balance between needed regulations to ensure scare public
investments in higher education are protected, without creating overly burdensome and
excessive requirements on institutions, increasing costs even further.
Limitations of the Study and Areas for Future Research
This study focused on the extent to which a cost disease affects higher education;
however, the analysis was limited by the datasets which were available. The higher
education price index was first developed in 1961, as a way to assess cost increases
specifically affecting higher education. The index was created is a composite,
aggregating both public and private non-profit institutions. However, public and private
non-profit institutions may face different cost structures. Beginning in 2001, the
CornmonFund Institute developed separate indices for public and private institutions.
Although the dataset is limited, future researchers could attempt to analyze the cost and
price behavior associated with each sector independently. This could enhance our
collective understanding of the cost pressures faced by the different sectors within higher
education.
As described in this chapter, the appropriate level of public support for higher
education hinges on an accurate assessment of the private versus social benefits
associated college attainment. This is a complex area of analysis, which involves some
Higher Education and the Cost Disease 142
subjective judgments. There have been a number of excellent studies conducted
analyzing the effects of higher education, both in the United States and abroad. However,
more research in this area may help to create a broader consensus on the impact of higher
education, and by extension, the appropriate level of public investment required to meet
federal and state policy objectives.
The discussion surrounding the relative distribution of private and social benefits
of higher education impacts the financing of higher education. Fruitful areas of research
include the appropriate funding mix between direct institutional subsidies for public
institutions and general need-based grants awarded to students attending both private and
public institutions, and the impact on access and equity through new mechanisms of
college financing such as income contingent loans. To be useful, this should include a
comparative analysis across various countries using different financing techniques. There
is much to learn kom the experience of other educational systems.
One of the conclusions of this study focuses on increasing the productivity
associated with colleges and universities. This includes increasing the effectiveness of
high school education to reduce the necessity of providing remedial coursework in
college, using technology to increase the opportunity for students to take courses, and
even serious exploration of a three-year degree program. More research needs to be
conducted in these areas to assist high schools, colleges and universities to more
effectively manage the academic enterprise, as they face rapidly rising costs with
relatively diminished resources.
One of the findings of this study suggests that the service sector is affected by a
cost disease, since the majority of government programs are service related, this has
Higher Education and the Cost Disease 143
broad implications for the area of public finance. While the service jobs are not devoid of
productivity gains, a general cost disease suggests that costs associated with government
programs will increase at rates faster than the general consumer price index. Moreover,
the United States is not the only country impacted; it seems this issue affects most OECD
countries as well.
Some of the costliest government supported programs including K-12 and higher
education as well as health care seem to be affected. Providing the appropriate level of
services may involve very contentious social issues, including the possible necessity of
increasing taxes, as well as limiting or delaying benefit programs previously thought of as
entitlements.
There has been some excellent research conducted on the impact of a cost disease
on the provision of government services. However, most of these are several decades old.
New research is needed to help guide public policy as we face very challenging resource
allocation decisions, not only among various government programs, but also associated
with the relative allocation of the economy between the public and private sectors.
Finally, the research revealed differing findings associated with the trajectory of
higher education costs as opposed to tuition sticker prices. While higher education costs
were found to be statistically non-significant compared with price increases associated
with NIPA Services, higher education tuition sticker prices were found to be statistically
significantly higher than price increases associated with the Service sector.
The analysis section revealed two of the most important causes, burden shifting
and tuition discounting. The analysis indicated that the practice of tuition discounting is
widespread and growing, involving both public and private non-profit institutions.
Higher Education and the Cost Disease 144
Moreover, the practice involves the widespread use of discounting to support merit
scholarships, unrelated to need.
While college costs are indeed rising faster than the CPI, reflecting a cost disease,
the widespread use of tuition discounting as a method to finance merit scholarships
suggests that non-profit colleges and universities face incentives related to the behavioral
motivations suggested in the literature review, including the quest for excellence, fueled
by competition for prestige.
This suggests that a more holistic model of cost and price behavior should be
developed, encompassing both the behavioral criteria suggested by such researchers such
as Bowen (1980), Winston (1996,2003), and (Goethals et al., 1999), which also integrates
revenue and cost components as well.
The researcher has attempted to create a more inclusive model below. Future
researchers may attempt to continue research along these lines, seeking to determine the
relative importance each of the variables in influencing cost and price behavior of non-
profit colleges and universities. This may increase our collective understanding of the
complex processes impacting not only cost and price issues, but also the micro-economic
variables impacting the decision-making behavior of colleges and universities. The model
is provided in Appendix H.
Higher Education and the Cost Disease 145
Conclusion
College costs have been rising faster than the inflation rate for nearly every year
since the Higher Education Price Index was created in 1961. A large body of research has
attempted determine the causes for rapidly rising college costs and prices, which can be
divided into two broad categories. First are researchers who see the non-profit structure
of most colleges and universities creating motivations involving the pursuit of excellence
and prestige leading to behavior which encourages high expenditures. Bowen (1980) best
articulated this point of view with his revenue theory of costs, in which he hypothesized
that colleges would raise all the money they can, and spend all the money they raised.
Alternatively, a growing body of literature assigned rapidly rising institutional
costs to a cost disease. This theory placed higher education cost increases within a
broader phenomenon affecting the entire service sector, suggesting that due to limitations
of leveraging technology, costs associated with the service sector tend to increase faster
than those in the manufacturing sector.
This research attempted to analyze the extent to which a cost disease could
explain higher education cost and price increases. This involved two distinct steps. First,
the researcher analyzed the subcomponents of the Higher Education Price Index to
determine the influence of personnel costs on higher education compared to other HEPI
subcomponents. Secondly, higher education costs and prices increases were compared
against three broad sectors: Durable Goods, Non-Durable Goods and Services, along with
the aggregate price index.
Higher Education and the Cost Disease 146
Personal costs were found to be a major cost driver within the components of the
higher education price index, and ANOVA Post-hoc analysis concluded that higher
education costs were statistically non-significant compared to price increases associated
with the aggregate Services index. This suggests that higher education costs are rising at
the same relative rate as other service industries, irrespective of whether the ownership
structure associated with particular service industries. This implies that cost increases
associated with higher education are unrelated to the non-profit status of the majority of
colleges and universities.
Moreover, price increases associated with non-durable goods were statistically
non-significant compared to increases associated with durable goods (a proxy used for
the manufacturing sector), while neither were statistically significantly different from
aggregate price increases.
However, price increases associated with Services were found to be significantly
higher than those associated with durable goods, suggesting that Service sector industries
face a general cost disease. Additionally, price increases associated tuition sticker prices
were found to be significantly higher than those associated with the Service sector.
Two main causes were assessed, including burden shifting and tuition
discounting. It does appear that government support for higher education has failed to
keep pace with increases in college costs on a per-capita student basis. However, it also
seems that the practice of tuition discounting is both widespread and growing; this
suggests that behavioral motivations such as the pursuit of excellence and competition for
prestige impact non-profit institutions of higher education. The researcher proposed a
new conceptual model for higher education cost and price behavior which encompasses
Higher Education and the Cost Disease 147
both behavioral as well as cost and revenue components to potentially explain higher
education cost and price increases.
The presence of a cost disease poses serious public policy challenges; it suggests
that the costs of government programs, which are primarily service-based, will increase
faster than general inflation. For higher education, it means that there will be a variety of
socially beneficial programs competing for public resources. Institutions of higher
education hold a privileged position in society; however, public support must be based on
the positive social benefits higher education provides to the broader society it serves.
These benefits are considerable, including enhanced productivity and output, a more
competitive workforce, a larger tax base, and broader civic responsibility.
However, the higher education community cannot take public support for granted;
faced with a cost disease, institutions must increase the learning productivity of their
processes, first articulated by Johnstone. This includes pursing objectives to increase
persistence and graduation, and finding ways to complete a degree in less time than
currently required.
A cost disease has serious implications for the way we finance higher education;
rapidly rising costs have created a growing debt burden, especially as government
support has failed to keep pace with higher education costs on a per-capita student basis.
We should explore new methods of financing which minimizes the risk of crushing
student debt after graduation, particularly through income contingent loans.
Since the social benefits of education accrue irrespective of whether a student
attends a private or public institution, we should reconsider the allocation of public
Higher Education and the Cost Disease 148
support between direct subsidies targeted to public institutions, and expanding need-
based aid programs for all students.
Higher education is still one of the primary mechanisms for ensuring equality of
opportunity; it is also vital to national prosperity and enhanced productivity in a rapidly
globalizing economy focused on knowledge creation. A cost disease threatens to create a
serious bottleneck in access to a college education, at a time when other countries are
expanding the knowledge base of their economy through investment in higher education.
Hopefully, the findings in this study can help reduce the contentious debate concerning
the causes of hgher education cost and price increases, and refocus efforts on ensuring
continued access to higher education to increase national prosperity.
Higher Education and the Cost Disease 149
References
1998 Amendments to the Higher Education Act of 1965, Pub. L. No. 105-244,105th
Congress (1998).
Adelman, C. (2008). The Bologna Club: What US higher education can learn from a
decade of European reconstruction: Institute for Higher Education Policy.
Adelman, C. (2009). The Bologna Process for US. Eyes: Re-learning Higher Education
in the Age of Convergence: Institute for Higher Education Policy.
Affordability in Higher Education Act, H.R. 3311 Sec 301, 108th Congress (2003).
Alexander, L. (2009, October 17,2009). The Three-Year Solution: How the reinvention
of higher education benefits parents, student, and schools. Newsweek.
Allen, E., & Seaman, J. (2007). Online nation: Five years ofgrowth in online learning.
Needman, MA: Sloan-C: The Sloan Consortium.
American Council on Higher Education. (2006). ACE fact sheet on higher education:
Pel1 grantfinding histoy (1976 to 2006). Washington, DC.
Archibald, R. B., & Feldman, D. H. (2008). Explaining increases in higher education
costs. The Journal ofHigher Education, 79(3), 268-295.
Barr, N. (1993). Alternative funding sources for higher education. The Economic Journal,
103(418), 718-728.
Basken, P. (2010, February, 25). Duncan defends cost and direction of education budget
to congress. Chronicle of Higher Education.
Baum, S., & Lapovsky, L. (2006). Tuition discounting: Notjust aprivate college
practice. New York: College Board.
Higher Education and the Cost Disease 150
Baum, S., Payea, K., & Steele, P. (2009). Trends in student aid 2009. Washington, D.C.:
The College Board.
Baumol, W. J. (1967). Macroeconomics of Unbalanced Growth: The Anatomy of Urban
Crisis. American Economic Review, 57(3), 415-426.
Baumol, W. J. (1993). Health care, education and the cost disease: A looming crisis for
public choice. Public Choice, 77(1), 17-28.
Baumol, W. J., Blackman Batey, S. A,, & Wolff, E. N. (1985). Unbalanced growth
revisited: Asymptotic stagnancy and new evidence. American Economic Review,
75(4), 806-817.
Baumol, W. J., &Blinder, A. S. (1985). Economics: Principles andpolicy (3rd ed.). San
Diego, CA: Harcourt Brace Jovanovich.
Baumol, W. J., & Bowen, W. G. (1965). On the performing arts: The anatomy of their
economic dilemma. American Economic Review, 50(2), 495-502.
Baumol, W. J., & Bowen, W. G. (1966). Performing arts - The economic dilemma. New
York: Twentieth Century Fund.
Blundell, R., Dearden, L., Goodman, A,, & Reed, H. (2000). The returns to higher
education in Britain: Evidence from a British cohort. The Economic Journal,
110(461), F82-F89.
Bowen, H. R. (1980). The costs of higher education: How much do colleges and
universities spendper student, and how much should they spend? San Francisco,
CA: Jossey-Bass Publishers.
Higher Education and the Cost Disease 15 1
Brennan, D. W. (2001). Study of college costs andprices, 1988-89 to 1997-98: Volume 2:
commissionedpapers, an essay on college costs. Retrieved May 20,2008. from
htto:/lnces.ed.eovlpubs2002/20021S8.pdf.
Brovender, S. (1974). On the economics of a university: Toward the determination of
marginal cost of teaching services. The Journal of Political Economy, 82(3), 657-
664.
Bureau of Economic Analysis. (2009). Concepts and Methods ofthe US . National
Income and Product Accounts: Chapter 5: Personal consumption expenditures.
Retrieved July 17,2009. from
http://www.bea.gov/nationaV~df/NIPAhandbookchS.pdf.
Camegie Commission on Higher Education. (1973). Higher Education: Who Pays? Who
Benefits? Who Should Pay? New York: McGraw-Hill Book Company.
Camevale, A. P., Smith, N., & Strohl, J. (2010). Help wanted: Projections ofjobs and
education requirements through 2018. Washington, D.C.: Georgetown
University.
Caroll, S. J., & Emre, E. (2009). The benefits to tanpayers from increases in students'
educational attainment. Santa Monica, CA: RAM) Corporation.
Chapman, B. (1997). Conceptual issues and the Australian experience with income
contingent loans for higher education. The Economic Journal, 107(442), 738-751.
Clotfelter, C. T. (1996). Buying the best: Cost escalation in elite higher education.
Princeton, NJ: National Bureau of Economic Research.
Clotfelter, C. T. (1999). The familiar but curious world of higher education: Introduction
to a symposium. Journal of Economic Perspectives, 13(1), 3-12.
Higher Education and the Cost Disease 152
College Access and Opportunity Act, H.R. 4283, Sec 131, 108th Congress (2003).
Common Core State Standards Initiative. (2010). Introduction to Common Core State
Standards. Washington, D.C.: Council of Chief State School Officers,
National Governors' Association Center for Best Practices.
CommonFund. (2004). College and University Higher Education Price Index. Wilton,
CT: CommonFund Institute.
Commonfund Institute. (2004). College and university higher education price index.
Wilton, CT: ComrnonFund Institute.
Commonfund Institute. (2005). CommonFund institute hepi ZOO5 update. Wilton, CT:
CommonFund Institute.
Congressional Budget Office. (1994). Issues in designing a federalprogram of income-
contingent student loans. Washington, D.C.: Congressional Budget Office.
Congressional Budget Office. (201 0). The Budget and economic outlook: Fiscal years
2010 to 2020. Retrieved June 7,2010. from
http://www.cbo.~ov/fipdocs/108xx/doc10871/~ontmatter.shtml.
Cunningham, A. F., Wellman, J. V., Clinedinst, M. E., Meriosotis, J. P., & Carroll, C. D.
(2001). Study of college costs andprices, 1988-89 to 1997-98. Washington: U.S.
Department of Education, Office of Educational Research and Improvement.
Department of Education. (2010). Repayment plans. Retrieved June 7,2010, from
htt~://www.direct.ed.eov/Re~a~Calc/dlindex2.html
Ehrenberg, R. G. (1999). Adam smith goes to college: An economist becomes an
academic administrator. Journal of Economic Perspectives, 13(1), 99-1 16.
Higher Education and the Cost Disease 153
Ehrenberg, R. G. (2000). Tuition rising: Why college costs so much. Cambridge, MA:
Harvard University Press.
Erslanger, S. (2010, May 24). Europeans fear crisis threatens liberal benefits. The New
York Times.
Fain, P. (2010, March 30,2010). Tuition-Discount Rate Hits Record High, Nacubo
Survey Finds. Chronicle of Higher Education.
Ferris, S. J., & West, E. G. (1996). The Cost disease and government growth:
Qualifications to Baumol. Public Choice, 89, 35-52.
Finn, C. E., Jr. (1978). Scholars, dollars and bureaucrats. Washington: The Brookings
Institution.
Goethals, G., Winston, G., & Zimmerman, D. (1999). Students educating students: The
emerging role ofpeer effects in higher education (No. DP-50). Williamstown,
MA: Williams College.
Gundlach, E., Wossmann, L., & Gmelin, J. (2001). The decline of schooling productivity
in OECD countries. The Economic Journal, 111(471), C135-C147.
Hansmann, H. (1980). The rationale for exempting nonprofit organizations ffom
corporate income tax. Yale Law Journal, 91, 54-100.
Hanson, K. H. (1993). The economics of independent higher education. In R. T. Ingram,
et al. (Ed.), Governing independent colleges and universities: A handbook for
trustees, chief executives and other campus leaders @p. 38-60). San Francisco,
CA: Jossey-Bass Publishers.
Hauptman, A. (1990). The tuition dilemma: Assessing new ways to pay for college.
Washington: The Brookings Institution.
Higher Education and the Cost Disease 154
Hauptman, A. (1997). Financing American higher education in the 1990's. New
Directions for Institutional Research, 24(1), 19-35.
Hebel, S. (2010, March 14). State cuts are pushing public colleges into peril. Chronicle of
Higher Education.
The Higher Education Act of 1965, Pub. L. No. 89-329,89th Congress (1965).
Higher Education Opportunity Act, H.R. 4137, Sec 801,llOth Congress (2007).
Highlights of Obama's Fiscal 201 1 Budget for Higher Education and Science. (2010,
February 23). Chronicle ofHigher Education.
Horn, L., Wei, C. C., Berker, A., & Carroll, C. D. (2002). What studentspay for college:
Changes in net price of college attendance between 1992-93 and 1999-2000.
Retrieved May 18,2008. from httu://nces.ed..gov/pubs2002/2002174.pdf.
Immenvahl, J., & Johnson, J. (2007). Squeezeplay: Howparents and thepublic look at
higher education today. San Jose, CA: National Center for Public Policy in
Higher Education.
James, E. (1978, Spring). Product mix and cost disaggregation: A reinterpretation of the
economics of higher education. Journal of Human Resources, 13, 157-186.
Johnstone, D. B. (1993). Learningproductivity: A new imperative for American higher
education.
Johnstone, D. B. (2001). Higher education and those "out-of-control costs". In P. G.
Altbach, P. J. Gumport & D. B. Johnstone (Eds.), In defense ofAmerican higher
education (pp. 144-178). Baltimore, MD: Johns Hopkins University Press.
Johnstone, D. B. (2003). The economics and politics of cost sharing in higher education:
Comparative perspectives. Economics of Education Review, 23(2004), 403-410.
Higher Education and the Cost Disease 155
Jones, D. P. (2001). Study of college costs andprices, 1988-89 to 1997-98: Volume 2:
Comissioned Papers, cost analysis and the formulation ofpublic policy. Retrieved
October, 30,2007. from httu://nces.ed.~ov/~ubs2002/2002158.pdf.
Kean, S. (2006, September 15). Report blames college practices for limiting access of
minority and low-income students. Chronicle of Higher Education.
Labi, A. (2009, April 8). Europe's higher education restructuring holds lessons for US.,
report says. The Chronicle of Higher Education.
Labi, A. (2010, May 25). Report assesses the recession's toll on higher education in
Europe. The Chronicle of Higher Education.
Lederman, D. (2007). Warning on tuition, shift on accreditation [Electronic Version].
Inside Higher Education, Nov 15,2007. Retrieved November 19,2007 from
httu://www.insidehi&ered.com/news/200711 111 Shea.
Marcuss, R. D., & Kane, R. E. (February, 2007). U.S. National income andproduct
statistics: Born of the great depression and world war II. Retrieved October 20,
2009. &om
htt~:Nwww.bea.~ov1scb/udE/2007/02%20feb~0207 history articleadf.
Massy, W. F. (2003). Honoring the trust: Quality and cost containment in higher
education. Bolton, MA: Anker Publishing Company, Inc.
Massy, W. F. (Ed.). (1996). Resource allocation in higher education. Ann Arbor, MI:
The University of Michigan Press.
McCully, C. P., & Teensma, T. D. (2008). Preview of the 2009 comprehensive revision of
the National Income and Product Accounts: New classzjication for personal
consumption expenditures. Retrieved. from.
Higher Education and the Cost Disease 157
Sewall, M. (2010, June 2,2010). New National Standards Seek to Make All Students
Ready for College. Chronicle of Higher Education.
Snower, D. J. (1 993). The future of the welfare state. The Economic Journal, 1 O3(418),
700-717.
Snyder, T. D., & Dillow, S. A. (2010a). Table 337. In Digest of Educational Statistics
(Ed.). Waslnngton, DC: National Center for Educational Statistics,.
Snyder, T. D., & Dillow, S. A. (2010b). Table 352. In Digest of Education Statistics
(Ed.). Washington, DC: National Center for Educational Statistics.
Snyder, T. D., & Dillow, S. A. (2010~). Table 353. In Digest of Educational Statistics
(Ed.). Washington, DC: National Center for Educational Statistics,.
Snyder, T. D., & Dillow, S. A. (2010d). Table 355. In Digest of Educational Statistics
(Ed.). Washington, DC: National Center for Educational Statistics,.
States in the red. (2010, May 14). The Wall Street Journal Digital Network.
Trachtenberg, S. J., & Kauvar, G. (2010, May 24). A degree in three. The New York
Times.
Vedder, R. (2007). Over invested and overpriced: American higher education today.
Washington, DC: Center for College Affordability and Productivity.
Wellman, J. V. (2001). Looking back, going forward: The Carnegie Commission tuition
policy. Washington, D.C.: Institute for Higher Education Policy.
Winston, G. C. (1 996). The economic structure of higher education: Subsidies, customer-
inputs, and hierarchy (No. DP-40). Williarnstown, MA: Williams College.
Winston, G. C. (1999). Subsidies, hierarchy, and peers: The awkward economics of
higher education. The Journal of Economic Perspectives, 13(1), 13-36.
Higher Education and the Cost Disease 158
Winston, G. C. (2003). Toward a theory of tuition: Prices, peer wages, and competition
in higher education (No. DP-65). Williamstown, MA: Williams College.
Winston, G. C., & Carbone, J. C. (2001). Saving, wealth, performance, and revenues in
us colleges and universities (No. DP-59). Williamstown, MA: Williams College.
Winston, G. C., Carbone, J. C., & Lewis, E. G. (1998). What's been happening to higher
education? A reference manual: 1986-87 to 1994-95. Williamstown, MA:
Williams College.
Winston, G. C., & Zirnmerman, D. J. (2000). Where is aggressiveprice competition
taking higher education (No. DP-56). Williamstown, MA: Williams College.
Wolanin, T. R. (2005). The higher education act reauthorization: Issues and prospects
[Electronic Version]. NEA 2005 Almanac of Higher Education, 39-5 1. Retrieved
May 15,2008 fkom httu:/lwww2.nea.org/he/healma2k5/images/a05o39.udf.
Zemsky, R., &Massy, W. F. (1995). Toward an understanding of our current
predicaments. Change, 27(6), 40-49.
Higher Education and the Cost Disease 159
Appendix A: Definitions *
Cost: Amount institutions spend to provide education and related services to students These are measured through expenditures.
Price: The amount students and their families are charged and what they pay for educational services. There are a number of different prices, including sticker price, price of attendance, and net price.
Sticker price: The tuition and fees charged by an institution. (also called the Published Price).
Total Price of Attendance: Tuition and fees (sticker price) that institutions charge students plus other expenses, including housing (room and board if a student is living on campus, or rent and other housing costs for students not living on campus). This is also called the Cost of Attendance.
Net Price: The amount students and their families pay after financial aid is subtracted from the total price of attendance.
Revenue: Current fund revenues which institutions receive kom a variety of sources, including tuition and fees, earnings from endowment income, government appropriations, government and private grants, and contracts, private gifts, the sale of educational services (e.g. dormitories and bookstores) and auxiliary enterprises.
Expenditures: Institutional spending for a variety of operating budget categories, including direct instruction, research, public service, academic support, student services, institutional support, operation and maintenance of plant, and scholarships and fellowships.
General Subsidy: The difference between the average price charged to students and the average cost to the institution for providing an education to a student, on a per-capita basis. Since institutions receive revenue from a wide variety of sources, both tuition and non-tuition based, almost all students receive a subsidy whether or not they attend private or public institutions, and whether or not they receive financial assistance. The general subsidy does not include an additional subsidy which some students receive in the form of scholarships and need-based aid. In many ways, institutional decisions about tuition and fees are also decisions about setting the level of the general subsidy.
Average Tuition: Institutions charge different categories of students different levels of tuition and fees. For example, there may be different fees and direct charges assessed for additional lab expenses, health services and exercise facilities. Moreover, many students
Higher Education and the Cost Disease 160
receive tuition dmcounts in the form of institutional aid, which results in net tuition prices which are lower than published tuition charges, or the 'sticker price".
* Appendix A; (Cunningham et al., 2001, p. 5).
Higher Education and the Cost Disease 161
Appendix B: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment
Appendix B: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment Dataset
Service Total Personal Admin Clerical Employees Compensation
25.4 26.5 27.8 29.1 30.5 32.3 34.3 36.6 39.2 42.1 44.8 47.1 49.8 52.8 56.3 60.0 63.5 67.6 72.4 78.4 85.8 93.5
Year CPI Index
Source: Higher Education Price Index, 2004 Update, pp 21-22, HEPl2009 Update
Personal Compensation Compensation Weight 74.8
Non HEPl Professional Professional Fringe Index Salaries Salaries Benefits Faculty
Appendix B: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment Dataset
Source: Complied HEPl2005 Update, pg. 3, HEPl2004 Update, pp. 21-22
Year CPI Index
Source: Higher Education Price Index, 2004 Update, pp 21-22, HEPl2009 Update
Personal Compensation Compensation Weight 74.8
Non HEPl Professional Professional Fringe Service Total Personal Index Salaries Salaries Benefits Faculty Admin Clerical Employees Compensation
Appendix 6: Higher Education Price Index. Personal Consumption and Contracted Supplies and Equipment Dataset 1 64
Source: Higher Education Price Index, 2004 Update, pp 21-22, HEPl2009 Update
Contracted Services,Supplies and Equipment Contract Services Weight 25.2
Total Supplies and Library Contracted
Year Misc Services Materials Equipment Acquisitions Utilities Services
Subindexes of Salaries of Professional Personnel Used for HEPl
Extension1 HEPl Public Administration1 Faculty Grad Asst Service lnst Services
Appendix 6: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment Dataset 165
Contracted Services,Supplies and Equipment Contract Services Weight 25.2
Total Supplies and Library Contracted
Year Misc Services Materials Equipment Acquisitions Utilities Services 1992 145.7 115.2 126.3 193.8 93.3 126.9 1993 149.5 113.2 128.6 203.4 94.7 129.4 1994 154.8 1 14.3 130.8 213.6 98.7 133.6 1995 158.0 115.7 133.5 220.2 96.8 135.3 1996 163.8 130.1 137.0 230.9 93.3 139.9 1997 167.3 128.6 139.3 253.4 106.1 147.2 1998 172.8 126.2 141.3 266.5 111.1 151.E 1999 177.0 123.2 143.3 282.1 100.5 150.8 2000 182.9 123.1 145.0 298.6 104.9 155.8 2001 199.8 131.8 147.3 317.4 169.9 172.2 2002 205.8 128.2 118.1 2003 209.5 132.2 157.6 2004 216.4 135.6 176.4 2005 222.7 145.5 200.2 2006 228.8 158.1 255.7 2007 238.3 165.3 220.6 2008 246.4 180.0 252.0
ubindexes of Salaries of Professional Personnel sed for HEPl
Extension1 €PI Public Administration1 acuity Grad Asst Sewice lnst Services
161.1 153.9 161.1 163.6 165.2 157.6 165.2 168.8 170.1 162.7 170.1 175.6 176.1 171.0 176.1 179.7 181.2 174.9 181.2 188.3 186.6 180.8 186.6 194.7 192.9 187.2 192.9 200.9 199.9 193.7 199.9 209.7 207.3 199.7 207.3 219.6 214.5 207.7 214.5 229.2 222.7 236.4 229.9 255.7 234.2 263.3 240.7 274.0 248.2 287.7 257.6 299.2 267.4 314.0
Source: Higher Education Price Index. 2004 Update. pp 21-22, HEPl2009 Update
Appendix B: Higher Education Price Index. Personal Consumption and Contracted Supplies and Equipment Dataset
Library Professional Year Personnel Salaries Tota
Subindexes of Salaries of Professional Personnel Used for the HEPl
HEPl HEPl HEPl Faculty Admin Benefit SupplyIEquip
Source: Higher Education Price Index, 2004 Update, pp 21-22, HEPl2009 Update
Appendix B: Higher Education Price Index, Personal Consumption and Contracted Supplies and Equipment Dataset
Library Professional Year Personnel Salaries Tota
1992 151.6 160.8 1993 155.5 165.0 1994 160.7 170.3 1995 167.1 176.1 1996 172.2 181.7 1997 176.6 187.2 1998 182.0 193.5 1999 188.8 200.7 2000 195.9 208.4 2001 198.6 215.e 2002 2003 2004 2005 2006 2007 2008
ubindexes of Salaries of Professional ersonnel Used for the HEPl
EPI HEPl HEPl aculty Admin Benefit SupplyIEquip
161.1 163.6 194.3 126.9 165.2 168.8 204.3 129.4 170.1 175.6 213.6 133.6 176.1 179.7 221.4 135.3 181.2 188.3 224.5 139.9 186.6 194.7 226.7 147.2 192.9 200.9 236.7 151.6 199.9 209.7 239.2 150.8 207.3 219.6 254.6 155.8 214.5 229.2 261.7 172.2 222.7 236.4 277.1 229.9 255.7 292.3 234.2 263.3 312.8 240.7 271 .O 327.2
343.7 360.8 374.2
Source: Higher Education Price Index, 2004 Update, pp 21-22, HEPl2009 Update
Higher Education and the Cost Disease 168
Appendix C: Consumer Price Index, Higher Education Price Index, and Major Subcomponents, 19613001
Appendix C: Higher Education Price Index, Personal Consumption, and Contracted Supplies and Equipment, 1961-2008, Including Contracted Services, Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 169
Personal Compensation Compensation Weight 74.8
Non HEPl Professional Professional Fringe
Year CPI Index Index Salaries Salaries Benefits Faculty
Total I
Source: Higher Education Price Index, 2004 Update, Higher Education Price Index, 2009 Update
Personal Service Compens
Admin Clerical Employees ation Misc Services
Appendix C: Higher Education Price Index, Personal Consumption, and Contracted Supplies and Equipment, 1961-2008, Including Contracted Services, Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 170
Source: Higher Education Price Index, 2004 Update, Higher Education Price Index. 2009 Update
Personal Compensation Compensation Weight 74.8
Total Non Personal
HEPl Professional Professional Fringe Service Compens Year CPl Index Index Salaries Salaries Benefits Faculty Admin Clerical Employees ation
1992 140.8 153.5 160.8 140.2 194.3 161.1 163.6 162.4 1993 145.2 157.9 165.0 144.2 204.3 165.2 168.8 167.6 1994 148.8 163.3 170.3 148.2 213.6 170.1 175.6 173.3 1995 153.2 168.1 176.1 152.5 221.4 176.1 179.7 179.1 1996 157.4 173.0 181.7 157.3 224.5 181.2 188.3 184.1 1997 161.9 178.4 187.2 162.1 226.7 186.6 194.7 189.0 1998 164.8 184.7 193.5 168.0 236.7 192.9 200.9 195.8 1999 167.6 189.1 200.7 174.1 239.2 199.9 209.7 202 .O 2000 172.5 196.9 208.4 180.4 254.6 207.3 219.6 210.8 2001 178.4 208.7 215.8 187.9 261.7 214.5 2292 197.7 182.6 218.1 2002 181.6 212.7 225.0 198.7 277.1 222.7 236.4 205.4 189.6 228.7 2003 185.5 223.5 234.3 203.8 292.3 229.4 255.7 21 1.1 193.9 238.1 2004 189.6 231.7 239.2 208.9 312.8 234.2 263.3 217.1 197.6 245.6 2005 195.3 240.8 246.4 214.1 327.2 240.7 274.0 223.4 201.4 253.7 2006 202.7 253.1 254.9 219.4 343.7 248.2 287.7 229.5 205.5 262.9 2007 208.0 260.3 264.7 227.5 260.8 257.6 299.2 237.7 213.6 256.9 2008 215.7 273.2 275.3 234.7 380.7 268.1 314.0 245.1 220.5 285.1
Misc Services
145.7 149.5 154.8 158.0 163.8 167.3 172.8 177.0 182.9 199.8 205.8 209.5 216.4 222.7 228.8 238.3 246.4
Appendix C: Higher Education Price Index, Personal Consumption, and Contracted Supplies and Equipment, 1961 -2008, Including Contracted Services. Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 171
Contracted Services,Supplies and Equipment Contract Services Weight 25.2
Subindices of Salaries and Professional Personnel Used for HEPl
Source: Higher Education Price Index, 2004 Update, Higher Education Price Index, 2009 Update
Supplies Total and Library Contracted
Year Materials Equipment Acquisitions Utilities Services
Extension1 Administra Professional HEPl Public tionllnst Library Salaries Faculty Grad Asst Service Services Personnel Total
Appendix C: Higher Education Price Index, Personal Consumption, and Contracted Supplies and Equipment, 1961-2008, Including Contracted Services, Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 172
Contracted Services,Supplies and Equipment Contract Services Weight 25.2
Source: Higher Education Price Index, 2004 Update, Higher Education Price Index, 2009 Update
Subindices of Salaries and Professional Personnel Used for HEPl
Supplies Total and Library Contracted
Year Materials Equipment Acquisitions Utilities Services 1992 11 5.2 126.3 193.8 93.3 126.9 1993 113.2 128.6 203.4 94.7 129.4 1994 114.3 130.8 213.6 98.7 133.6 1995 115.7 133.5 220.2 96.8 135.3 1996 130.1 137.0 230.9 93.3 139.9 1997 128.6 139.3 253.4 106.1 147.2 1998 126.2 141.3 266.5 111.1 151.6 1999 123.2 143.3 282.1 100.5 150.8 2000 123.1 145.0 298.6 104.9 155.8 2001 131.8 147.3 317.4 169.9 172.2 2002 128.2 118.1 169.2 2003 132.2 157.6 179.1 2004 135.6 176.4 187.1 2005 145.5 200.2 197.8 2006 158.1 255.7 214.5 2007 165.3 220.6 216.4 2008 180.0 252.0 230.9
Extension1 Administra Professional HEPl Public tionllnst Library Salaries Faculty Grad Asst Service Services Personnel Total
161.1 153.9 161.1 163.6 151.6 160.8 165.2 157.6 165.2 168.8 155.5 165.0 170.1 162.7 170.1 175.6 160.7 170.3 176.1 171.0 176.1 179.7 167.1 176.1 181.2 174.9 181.2 188.3 172.2 181.7 186.6 180.8 186.6 194.7 176.6 187.2 192.9 187.2 192.9 200.9 182.0 193.5 199.9 193.7 199.9 209.7 188.8 200.7 207.3 199.7 207.3 219.6 195.9 208.4 214.5 207.7 214.5 229.2 198.6 215.8 222.7 236.4 229.9 255.7 234.2 263.3 240.7 274.0 248.2 287.7 257.6 299.2 267.4 314.0
Appendix C: Higher Education Price Index, Personal Consumption, and Contracted Supplies and Equipment, 1961-2008, Including Contracted Services. Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 173
Subindexes of Salaries of Professional Personnel Used for HEPl
HEPl HEPl HEPl supply/ Year Faculty Admin Benefit Equip
Source: Higher Education Price Index, 2004 Update, Higher Education Price Index, 2009 Update
Appendix C: Higher Education Price Index. Personal Consumption, and Contracted Supplies and Equipment, 1961-2008, Including Contracted Services, Supplies and Equipment, and Professional and Non-Professional Services from 2002-2008 174
Subindexes of Salaries of Professional Personnel Used for HEPl
HEPl HEPl HEPl supply/ Year Faculty Admin Benefit Equip
1992 161.1 163.6 194.3 126.9 1993 165.2 168.8 204.3 129.4 1994 170.1 175.6 213.6 133.6 1995 176.1 179.7 221.4 135.3 1996 181.2 188.3 224.5 139.9 1997 186.6 194.7 226.7 147.2 1998 192.9 200.9 236.7 151.6 1999 199.9 209.7 239.2 150.8 2000 207.3 219.6 254.6 155.8 2001 214.5 229.2 261.7 172.2 2002 222.7 236.4 277.1 2003 229.9 255.7 292.3 2004 234.2 263.3 312.8 2005 240.7 271 .O 327.2 2006 343.7 2007 360.8 2008 374.2
Source: Higher Education Price Index. 2004 Update. Higher Education Price Index, 2009 Update
Higher Education and the Cost Disease 175
Appendix D: Consumer Price Index, Higher Education Price Index, and Major Subcomponents, 1961 - 2001, Reindexed, 1961 = 100
Appendix D: Reindexed HEPl lndex and Subcomponents
Personal Compensation Compensation Weight 74.8
Non HEPl Professional Professional Fringe
Year CPI Index Index Salaries Salaries Benefits Faculty Admin Service Total Personal Misc
Clerical Employees Compensation Services 100.0
Source: Higher Education Price Index. 2004 Update. Higher Education Price Index, 2009 Update, CommonFund Institute. Reindexed by author.
Appendix D: Reindexed HEPl lndex and Subcomponents 1961 = 100
Personal Compensation Compensation Weight 74.8
Non HEPl Professional Professional Fringe Service Total Personal Misc
Year CPI Index Index Salaries Salaries Benefits Faculty Admin Clerical Employees Compensation Services
Source: Higher Education Price Index. 2004 Update. Higher Education Price Index, 2009 Update, CommonFund Institute. Reindexed by author.
Appendix D: Reindexed HEPl Index and Subcomponents 1961 = 100
Contracted Services,Supplies and Equipment Contract Services Weight 25.2 Subindexes of Salaries of Professional Personnel Used for the HEPl
Supplies Extension1 and Library Total Contracted HEPl Public Administration Library Professional
Year Materials Equi~ment Acauisitions Utilities Services Facultv Grad Asst Service llnst Services Personnel Salaries
Source: Higher Education Price Index. 2004 Update, Higher Education Price Index, 2009 Update, ComrnonFund Institute. Reindexed by author.
Appendix D: Reindexed HEPl Index and Subcomponents 1961 = 100
Contracted Services,Supplies and Equipment Contract Services Weight 25.2 Subindexes of Salaries of Professional Personnel Used for the HEPl
Supplies Extension1 and Library Total Contracted HEPl Public Administration Library Professional
Year Materials Equipment Acquisitions Utilities Services Faculty Grad Asst Service /Inst Services Personnel Salaries 1992 343.9 360.9 1300.7 594.3 486.2 161.1 153.9 161.1 163.6 151.6 160.8 1993 337.9 367.4 1365.1 603.2 495.8 165.2 157.6 165.2 168.8 155.5 165.0 1994 341.2 373.7 1433.6 628.7 511.9 170.1 162.7 170.1 175.6 160.7 1 70.3 1995 345.4 381.4 1477.9 616.6 518.4 176.1 171.0 176.1 179.7 167.1 176.1 1996 388.4 391.4 1549.7 594.3 536.0 181.2 174.9 181.2 188.3 172.2 181.7 1997 383.9 398.0 1700.7 675.8 564.0 186.6 180.8 186.6 194.7 176.6 187.2 1998 376.7 403.7 1788.6 707.6 580.8 192.9 187.2 192.9 200.9 182.0 193.5 1999 367.8 409.4 1893.3 640.1 577.8 199.9 193.7 199.9 209.7 188.8 200.7 2000 367.5 414.3 2004.0 668.2 596.9 207.3 199.7 207.3 21 9.6 195.9 208.4 2001 393.4 420.9 2130.2 1082.2 659.8 214.5 207.7 214.5 229.2 198.6 215.8 2002 382.7 752.2 648.4 222.7 236.4 2003 394.6 1003.8 686.2 229.9 255.7 2004 404.8 1123.6 716.9 234.2 263.3 2005 434.3 1275.2 757.9 240.7 274.0 2006 471.9 1628.7 821.9 248.2 287.7 2007 493.4 1405.1 829.1 257.6 299.2 2008 537.3 1605.1 884.8 267.4 314.0
Source: Higher Education Price Index, 2004 Update. Higher Education Price Index, 2009 Update. CommonFund Institute. Reindexed by author,
Appendix D: Reindexed HEPl Index and Subcomponents 1961 = 100
Subindices of HEPl Salaries and Professional Pesonnel
HEPl HEPl HEPl Year Faculty Admin Benefit
1961 1962 1963 1964 1965 I966 1967 1968 I969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 100.0 100 .O 100.0 1984 104.7 104.5 108.3 1985 111.6 110.7 117.7 1986 118.4 117.7 127.7 1987 125.4 124.8 137.4 1988 131.6 129.9 147.2 1989 139.2 139.3 158.8 1990 147.7 150.6 171.4 1991 155.7 159.1 184.3
Source: Higher Education Price Index. 2004 Update, Higher Education Price Index, 2009 Update, ComrnonFund Institute. Reindexed by author.
Subindices of HEPl Salaries and Professional Pesonnel
HEPl HEPl HEPl Year Faculty Admin Benefit
Appendix D: Reindexed HEPl Index and Subcomponents 1961 = 100
Source: Higher Education Price Index. 2004 Update, Higher Education Price Index, 2009 Update, CommonFund Institute. Reindexed by author.
Higher Education and the Cost Disease 182
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index, 1929 - 2008
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Line I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Item Personal consumption expenditures
Goods Durable goods Motor vehicles and parts New motor vehicles (55) Net purchases of used motor vehicles (56) Motor vehicle parts and accessories (58)
Furnishings and durable household equipment Furniture and furnishings (parts of 31 and 32) Household appliances (part of 33) Glassware, tableware, and household utensils (34) Tools and equipment for house and garden (35)
Recreational goods and vehicles Video, audio, photographic, and information processing equipment and media (75, 76, and part of 93) Sporting equipment, supplies, guns, and ammunition (part of 80) Sports and recreational vehicles (79) Recreational books (part of 90) Musical instruments (part of 80)
Other durable goods Jewelry and watches (part of 119) Therapeutic appliances and equipment (42) Educational books (96) Luggage and similar personal items (part of 119) Telephone and facsimile equipment (67)
Nondurable goods Food and beverages purchased for off-premises consumption Food and nonalcoholic beverages purchased for off-premises consumption (4) Alcoholic beverages purchased for off-premises consumption (5) Food produced and consumed on farms (6)
Clothing and footwear Garments Women's and girls' clothing (1 0) Men's and boys' clothing (1 1) Children's and infants' clothing (12)
Other clothing materials and footwear (13 and 17) Gasoline and other energy goods
Item Code DPCERG3 DGDSRG3 DDURRG3 DMOTRG3 DNMVRG3 DNPVRG3 DMVPRG3 DFDHRG3 DFFFRG3 DAPPRG3 DUTERG3 DTOORG3 DREQRG3 DVAPRG3 DSPGRG3 DWHLRG3 DRBKRG3 DMSCRG3 DODGRG3 DJRYRG3 DTAERG3 DEBKRG3 DLUGRG3 DTCERG3 DNDGRG3 DFXARG3 DTFDRG3 DAOPRG3 DFFDRG3 DCLORG3 DGARRG3 DWGCRG3 DMBCRG3 DCICRG3 DOCCRG3 DGOERG3
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index 184
Line 1 37 38 39 40 4 1 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
ltem Personal consumption expenditures Motor vehicle fuels, lubricants, and fluids (59) Fuel oil and other fuels (29)
Other nondurable goods Pharmaceutical and other medical products (40 and 41) Recreational items (parts of 80,92, and 93) Household supplies (parts of 32 and 36) Personal care products (part of 118) Tobacco (127) Magazines, newspapers, and stationery (part of 90) Net expenditures abroad by US. residents (131)
Services Household consumption expenditures (for services) Housing and utilities
Housing Rental of tenant-occupied nonfarm housing (20) Imputed rental of owner-occupied nonfarm housing (21) Rental value of farm dwellings (22) Group housing (23)
Household utilities Water supply and sanitation (25) Electricity and gas Electricity (27) Natural gas (28)
Health care Outpatient services Physician services (44) Dental services (45) Paramedical services (46)
Hospital and nursing home services Hospitals (51) Nursing homes (52)
Transportation services Motor vehicle services Motor vehicle maintenance and repair (60) Other motor vehicle services (61)
ltem Code DPCERG3 DMFLRG3 DFULRG3 DONGRG3 DPHMRG3 DREIRG3 DHOURG3 DOPCRG3 DTOBRG3 DNEWRG3 2222223 DSERRG3 DHCERG3 DHUTRG3 DHSGRG3 DTENRG3 DOWNRG3 DFARRG3 DGRHRG3 DUTLRG3 DWRSRG3 DELGRG3 DELCRG3 DGHERG3 DHLCRG3 DOUTRG3 DPHYRG3 DDENRG3 DPMSRG3 DHPNRG3 DHSPRG3 DNRSRG3 DTRSRG3 DMVSRG3 DVMRRG3 DOVSRG3
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 185
Line 1
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 l o 4 105 lo6
ltem Personal consumption expenditures Public transportation Ground transportation (63) Air transportation (64) Water transportation (65)
Recreation services Membership clubs, sports centers, parks, theaters, and museums (82) Audio-video, photographic, and information processing equipment services (parts of 77 and 93) Gambling (91) Other recreational services (81, 94, and part of 92)
Food services and accommodations Food services Purchased meals and beverages (1 02) Food furnished to employees (induding military) (103)
Accommodations (1 04) Financial services and insurance
Financial services Financial services furnished without payment (107) Financial service charges, fees, and commissions (108)
Insurance Life insurance (1 10) Net household insurance (1 11) Net health insurance (1 12) Net motor vehicle and other transportation insurance (116)
Other services Communication Telecommunication services (71) Postal and delivery services (68) Internet access (72)
Education services Higher education (97) Nursery, elementary, and secondary schools (98) Commercial and vocational schools (99)
Professional and other services (121) Personal care and dothing services (14 and parts of 17 and 118) Social services and religious activities (120)
ltem Code DPCERG3 DPUBRG3 DGRDRG3 DAlTRG3 DWATRG3 DRCARG3 DRLSRG3 DAVPRG3 DGAMRG3 DOTRRG3 DFSARG3 DFSERG3 DPMBRG3 DFOORG3 DACCRG3 DIFSRG3 DFNLRG3 DIMPRG3 DOFIRG3 DINSRG3 DLIFRG3 DFINRG3 DHINRG3 DTINRG3 DOTSRG3 DCOMRG3 DTCSRG3 DPSSRG3 DINTRG3 DTEDRG3 DHEDRG3 DNEHRG3 DVEDRG3 DPRSRG3 DPERRG3 DSOCRG3
Source Bureau of Economlc Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
lo7 lo8 109 110 111 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
ltem Personal consumption expenditures Household maintenance (parts of 31, 33, and 36) Net foreign travel Foreign travel by U.S. residents (129) Less: Expenditures in the United States by nonresidents (130)
Final consumption expenditures of nonprofit institutions sewing households (NPISHs) \l\ Gross output of nonprofit institutions (133) \2\ Less: Receipts from sales of goods and services by nonprofit institutions (134) \3\
ltem Code DPCERG3 DHHMRG3 z222223 DFTRRG3 DEXFRG3 DNPIRG3 DNPERG3 DNPSRG3
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National lncome and Product Accounts. Personal Consumption Index, Table 2.4
Line I 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 lo4 105 lo6
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line I
107 lo8 109 110 Ill 112 113
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index 190
Source Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line 1
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 lo3 l o 4 105 l o 6
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
107 lo8 109 110 111 112 113
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Appendix E: National Income and Product Accounts, Personal Consurn~tion Ex~enditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 7 1
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 lo3 104 lo5 lo6
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4
Line 1 107 lo8 lo9 110 Ill 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index. Table 2.4
Line 1 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 7 1
Appendix E: National Income and Product Accounts, Personal Consumption Ex~enditures Index
Source Bureau of Economic Analysis. National lncome and Product Accounts. Personal Consumption Index, Table 2.4
Line I 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106
Appendix E: National Income and Product Accounts, Personal Consum~tion Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
l o7 108 l o9 110 111 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 202
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 203
Line I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 l o 6
Appendix E: National Income and Product Accounts, Personal Consum~tion Exwnditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
107 lo8 lo9 110 Ill 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 206
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 207
Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line I 37 38 39 40 4 1 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index, Table 2.4
Line 1 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 lo4 105 lo6
Appendix E: National Income and Product Accounts. Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index. Table 2.4
Line 1
107 108 109 110 111 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index 210
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Line I 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Source Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index. Table 2.4
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Line 1
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 l o 3 l o 4 105 l o6
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4
Line 1
107 l o 8 109 110 Ill 112 113
Appendix E: National Income and Product Accounts, Personal Consumption Expenditures Index
Source Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4
Higher Education and the Cost Disease 215
Appendix F: National Income and Product Accounts, Reinexed, Base Year 1961
Appendix F: National Income and Product Accounts. Personal Consumption Index, Reindexed Base Year 1961 = 100
Line Item 1 Personal consumption expenditures 2 Goods 3 Durable goods 4 Motor vehicles and parts 5 New motor vehicles (55) 6 Net purchases of used motor vehicles (56) 7 Motor vehicle parts and accessories (58) 8 Furnishings and durable household equipment 9 Furniture and furnishings (parts of 31 and 32)
10 Household appliances (part of 33) I 1 Glassware, tableware, and household utensils (34) 12 Tools and equipment for house and garden (35) 13 Recreational goods and vehicles
Video, audio, photographic, and information processing equipment and media (75, 76, and part of 14 93) 15 Sporting equipment, supplies, guns, and ammunition (part of 80) 16 Sports and recreational vehicles (79) 17 Recreational books (part of 90) 18 Musical instruments (part of 80) 19 Other durable goods 20 Jewelry and watches (part of 11 9) 21 Therapeutic appliances and equipment (42) 22 Educational books (96) 23 Luggage and similar personal items (part of 119) 24 Telephone and facsimile equipment (67) . . . . 25 on durable goods 26 Food and beverages purchased for off-premises consumption 27 Food and nonalcoholic beverages purchased for off-premises consumption (4) 28 Alcoholic beverages purchased for off-premises consumption (5) 29 Food produced and consumed on farms (6) 30 Clothing and footwear 31 Garments 32 Women's and girls' clothing (10) 33 Men's and boys' clothing (1 1 ) 34 Children's and infants' clothing (12) 35 Other clothing materials and footwear ( I3 and 17)
Category
Durable Durable Durable Durable Durable Durable Durable Durable Durable Durable Durable
Durable Durable Durable Durable Durable Durable Durable Durable Durable Durable Durable
ltem Code DPCERG3 DGDSRG3 DDURRG3 DMOTRG3 DNMVRG3 DNPVRG3 DMVPRG3 DFDHRG3 DFFFRG3 DAPPRG3 DUTERG3 DTOORG3 DREQRG3
DVAPRG3 DSPGRG3 DWHLRG3 DRBKRG3 DMSCRG3 DODGRG3 DJRYRG3 DTAERG3 DEBKRG3 DLUGRG3 DTCERG3
NonDurable DNDGRG3 NonDurable DFXARG3 NonDurable DTFDRG3 NonDurable DAOPRG3 NonDurable DFFDRG3 NonDurable DCLORG3 NonDurable DGARRG3 NonDurable DWGCRG3 NonDurable DMBCRG3 NonDurable DCICRG3 NonDurable DOCCRG3
Source: Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts. Personal Consumption Index. Reindexed Base Year 1961 = 100
Line Item 1 Personal consumption expenditures
36 Gasoline and other energy goods 37 Motor vehicle fuels, lubricants, and fluids (59) 38 Fuel oil and other fuels (29) 39 Other nondurable goods 40 Pharmaceutical and other medical products (40 and 41) 41 Recreational items (parts of 80, 92, and 93) 42 Household supplies (parts of 32 and 36) 43 Personal care products (part of 11 8) 44 Tobacco (127) 45 Magazines, newspapers, and stationery (part of 90) 47 Services 48 Household consumption expenditures (for services)
Housing and utilities Housing Rental of tenant-occupied nonfarm housing (20) Imputed rental of owner-occupied nonfarm housing (21) Rental value of farm dwellings (22) Group housing (23)
Household utilities Water supply and sanitation (25) Electricity and gas Electricity (27) Natural gas (28)
Health care Outpatient services Physician services (44) Dental services (45) Paramedical services (46)
Hospital and nursing home services Hospitals (51) Nursing homes (52)
Transportation services Motor vehicle services Motor vehicle maintenance and repair (60) Other motor vehicle services (61)
Category ltem Code DPCERG3
NonDurable DGOERG3 NonDurable DMFLRG3 NonDurable DFULRG3 NonDurable DONGRG3 NonDurable DPHMRG3 NonDurable DREIRG3 NonDurable DHOURG3 NonDurable DOPCRG3 NonDurable DTOBRG3 NonDurable DNEWRG3 Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 218
Line Item 1 Personal consumption expenditures
72 Public transportation 73 Ground transportation (63) 74 Air transportation (64) 75 Water transportation (65) 76 Recreation services 77 Membership clubs, sports centers, parks, theaters, and museums (82)
Audio-video, photographic, and information processing equipment services (parts of 77 and 93) Gambling (91) Other recreational services (81.94, and part of 92)
Food services and accommodations Food services Purchased meals and beverages (102) Food furnished to employees (including military) (103)
Accommodations (1 04) Financial services and insurance Financial services Financial services furnished without payment (107) Financial service charges, fees, and commissions (108)
Insurance Life insurance (1 10) Net household insurance (1 11) Net health insurance (1 12) Net motor vehicle and other transportation insurance (1 16)
Other services Communication Telecommunication services (71) Postal and delivery services (68)
Education services Higher education (97) Nursery, elementary, and secondary schools (98) Commercial and vocational schools (99)
Professional and other services (121) Personal care and clothing services (14 and parts of 17 and 118) Social services and religious activities (120)
Category ltem Code DPCERG3
Services DPUBRG3 Services DGRDRG3 Services DAITRG3 Services DWATRG3 Services DRCARG3 Services DRLSRG3
Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services Services
Source: Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 219
Line Item 1 Personal consumption expenditures
107 Household maintenance (parts of 31. 33, and 36) 109 Foreign travel by US. residents (129) 110 Less: Expenditures in the United States by nonresidents (130) 1 11 Final consumption expenditures of nonprofit institutions serving households (NPISHs) \ 112 Gross output of nonprofit institutions (133) \2\ 113 Less: Receipts from sales of goods and services by nonprofit institutions (134) \3\
Category ltem Code DPCERG3
Services DHHMRG3 Services DFTRRG3 Services DEXFRG3
,1 \ Services DNPIRG3 Services DNPERG3 Services DNPSRG3
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts. Personal Consumption Index. Reindexed Base Year 1961 = 100 220
Line 1 2 3 4 5 6 7 8 9
10 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author
Appendix F: National Income and Product Accounts. Personal Consumption Index. Reindexed Base Year 1961 = 100 221
Line 1 36 37 38 39 40 41 42 43 44 45 47 48 49 50 51 52 53 54 55 56 57 58
Source: Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index. Reindexed Base Year 1961 = 100 222
Line 1
72 73 74 75 76 77
78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
100 101 102 lo3 lo4 l o5 lo6
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4. Reindexed by author
Line 1
107 109 110 11 1 112 113
Appendix F: National Income and Product Accounts. Personal Consumption Index. Reindexed Base Year 1961 = 100 223
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index. Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index. Reindexed Base Year 1961 = 100 224
Line 1 2 3 4 5 6 7 8 9
10 11 12 13
14 15 16 17 18 19 20 2 1 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Source: Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 225
Line 1
36 37 38 39 40 41 42 43 44 45 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 66 69 70 71
Source: Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author
Appendix F: National Income and Product Accounts. Personal Consumption Index, Reindexed Base Year 1961 = 100 226
Line 1
72 73 74 75 76 77
78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
100 101 102 103 104 l o5 l o6
Source: Bureau of Economic Analysis, National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Line 1
107 109 110 111 112 113
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 227
Source: Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author
Appendix F: National Income and Product Accounts, Personal Consumption Index. Reindexed Base Year 1961 = 100 228
Line 1 2 3 4 5 6 7 8 9
$0 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index. Reindexed Base Year 1961 = 100 229
Line 1
36 37 38 39 40 41 42 43 44 45 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author
Appendix F: National Income and Product Accounts. Personal Consumption Index, Reindexed Base Year 1961 = 100 230
Line 1 72 73 74 75 76 77
Source: Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Line 1
lo7 109 110 111 112 113
Appendix F: National Income and Product Accounts. Personal Consumption Index. Reindexed Base Year 1961 = 100 231
Source: Bureau of Economic Analysis. National Income and Product Accounts. Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts. Personal Consumption Index, Reindexed Base Year 1961 = 100 232
Line 1 2 3 4 5 6 7 8 9
10 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Source: Bureau of Economic Analysis. National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100
Line 1
36 37 38 39 40 41 42 43 44 45 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 234
Line 1
72 73 74 75 76 77
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Line 1
l o 7 109 110 111 112 1 I 3
Appendix F: National Income and Product Accounts, Personal Consumption Index, Reindexed Base Year 1961 = 100 235
Source: Bureau of Economic Analysis, National Income and Product Accounts, Personal Consumption Index, Table 2.4. Reindexed by author.
Higher Education and the Cost Disease 236
Appendix G: Research Database Used to Conduct ANOVA Analysis
Appendix G: Research Database Used to Conduct ANOVA Analysis Durable Non Durable
Year All Goods Goods Goods Services Higher Education HEPI Index 1961 100.00 100.00 100.00 100.00 100.00 100.00
Source: Hiaher Education Price Index. 2004 U~date. Hioher Education Price Index. - - 2009 Update. National Income and Product Accounts. Personal Consumption Index, Table 2.4. Re~ndexed by author, Base Year 1961 = I00
Higher Education and the Cost Disease 238
Appendix H: Future Research: Creating a Model of Cost and Price Escalation in Higher Education
This research has focused on the extent to which cost and price escalation in
higher education can be explained by the presence of a cost disease. The research results
suggest that higher education costs are statistically not different from price increases
associated with the service sector, however, tuition sticker prices were found to be
significantly higher than those associated with the services. While the failure of
government support to increase at the rate of college costs on per-capita student basis is a
cause, this analysis also highlighted the widespread use of tuition discounting as well.
This suggests that a more holistic model be developed to explain cost and price behavior
associated with non-profit institutions of higher education, one which reflects the impact
of college costs and revenue, but also encompasses behavioral factors associated with the
pursuit of excellence and prestige and the competition which these factors may engender.
The model detailed below provides a conceptual framework to holistically view these
factors. Future researchers may attempt to assess the part each plays in impacting higher
education cost and price increases.
A Model of Cost and Price Escalation in Higher Education
The model detailed below provides an initial exploration of the interactions
between the market structure associated with higher education and its impact on cost and
price escalation. There are four major components within the model: a revenue
subsystem, an expenditure subsystem, a derived student subsidy system, and a quality
market composed of both students and institutions.
Higher Education and the Cost Disease 239
Revenue Subsystem
Institutions receive revenue from four potential sources: government, endowment
income dedicated to current operations, current gifts, as well as tuition revenue. Public
institutions receive a relatively greater share of their revenues from government, while
certain private institutions have substantial endowments which they use to fund current
operations. Depending on the amount of revenue provided by government or gifts,
institutions will rely on tuition to a greater or lesser extent to meet their revenue
requirements. The total amount of government support, endowment, or current gifts can
change from year to year based on increased or decreased government allocations, or
relative changes in the amount and return on investment for institutional gifts.
In order to make meaningful comparisons between institutions, we must create a
standard measure which accounts for the variability in the populations of students
attending different institutions, as well as different amount of their revenues and
expenditures. Thus, revenue and expenditures must be assessed on a per-capita basis.
(See section on Operationalizing Variables for more detail). Since institutions vary in the
number of full and part-time students they serve, the number of undergraduates should be
counted on a full-time equivalency basis. This also enables us to create equivalent
comparative revenue and expenditure assessments, through the addition of per-capita
revenue and per-capita expenditure variables.
Expenditure Subsystem
Expenditures are composed of personnel costs, non-personnel costs, and
institutional non-endowed aid funded from operating revenue which is not supported by
other sources of gift aid. This component of institutional non-endowed aid can be viewed
Higher Education and the Cost Disease 240
as the increasing reliance and importance of tuition discounting used by colleges and
universities as a way to tailor their institutional profile to increases certain indices of
student quality among their enrolled population of students.
The total amount spent for personnel, non-personnel items and institutional merit
aid items associated with undergraduate instruction yields the total UG Instructional and
General Educational Costs. When assessed against the number of full-time equivalent
undergraduates, we are able to derive Per Capita Educational and General Costs.
As detailed in the literature review, one of the potential issues affecting annual
tuition increases in higher education are the rate of cost increases associated with the
production function of colleges and universities (Baumol, 1967, 1993; Baumol & Bowen,
1966; Bowen, 1980; Clotfelter, 1996, 1999; Ehrenberg, 1999,2000).
These can be seen as intrinsic costs associated with the purchased labor and non-
labor inputs used in producing educational services provided in undergraduate education.
The model takes these into account with two variables - a Personnel Cost Increase Rate
and Non-Personnel Cost increase Rate, which impact Personnel and Non-Personnel Costs
for the upcoming academic year.
The Market for Higher Education: Quality Markets and Student Subsidies
However, there is great variability in the amount institutions spend in the
production of educational services, as well as the tuition they charge (Bowen, 1980;
Winston, 1996). Further, the pressures on cost and price seem to be connected to the
market structure and prestige hierarchy associated with institutions of higher education.
Any model attempting to explain both the variability in tuition prices and spending by
institutions along with the cost and price pressures they face must account for the
Higher Education and the Cost Disease 241
structure of the higher education market which influences their behavior. (Rothschild &
White, 1995; Winston, 1999,2003; Winston & Zimmerman, 2000). As described by
Rothschild (1995) and Winston (1996, 1999), the market for higher education is
somewhat unique; it utilizes a customer input technology in which the presence of
particular students influences the quality of the educational experience of other students.
Under these conditions, the student is simultaneously both a customer and an input into
the educational process.
This is reflected in the "Quality Market" subsystem used in the model. The
quality market is composed of both institutions (Tier Institutional Quality) and students
(Student Quality). Institutional selectivity helps define both institutional quality as well as
the average quality indicators of those applicants who are accepted and eventually enroll
at the institution.
However, institutional quality is partially defined through competition with and in
comparison to other institutions, particularly those considered within one's peer group
(Clotfelter, 1999; Ehrenberg, 2000; Massy, 2003; Winston, 1999,2003; Winston &
Zimmerman, 2000). These competitive pressures, particularly on spending, are reflected
in the influence of the Average Peer Group Per Capita Subsidy and average Peer Group
Instructional and Educational Expenditures on institutional quality.
Student Subsidy Subsystem
The competition between institutions for student quality is heavily influenced by
the resources available to institutions. Institutions which are able to provide significant
donative resources purposely create an excess demand queue for their product, since they
Higher Education and the Cost Disease 242
are able to provide a substantial return on a student's tuition investment. They use this
excess demand queue to select those students whom they feel will contribute to peer
quality. They are thus able to create a high quality academic program by ensuring they
remain selective in choosing their student body (Winston, 1999).
This critical link between the donative resources provided by institutions and
measures of institutional and student quality is encapsulated in the Student Subsidies
subsystem. Tier Institutional Quality and Institutional Selectivity are heavily influenced
by the Per Capita Non-Tuition General Subsidy made possible through non-tuition based
revenue available to an institution.
In addition to donative resources used to provide a general subsidy for all
students, institutions can utilize a portion of their donative non-tuition revenues as well as
aid based on tuition discounting to provide individual student subsidies. This is
increasingly being used by institutions which do not have large endowments as the most
elite institutions, often in an attempt to provide attractive, targeted financial aid packages
to entice highly desirable students to enroll. This is reflected in the model through the
inclusion of the Per Capita Individual Subsidy influencing matriculation decisions, based
on indices of student quality.
Tuition Sticker Prices and Tuition Increases
Tuition sticker prices and tuition increases can be viewed as the residual effects of
the strange market associated with higher education; tuition increases are generated
through differences between the total Per Capita subsidy available to undergraduate full-
Higher Education and the Cost Disease 243
time equivalent students, and the Per Capita Educational and General Costs expended on
a full-time equivalency basis.
However, it is a market where educational services are consistently sold at a loss,
since institutions subsidize costs to the extent that their donative resources allow. The
competition for student excellence creates unusual cost and price pressures on
institutions, which I hope to explore in greater detail utilizing systems dynamics
modeling. A visual representation of the model is displayed in Figure 1.
Higher Education and the Cost Disease 244
Figure 1: Model of Cost and Price Escalation in Higher Education
Higher Education and the Cost Disease 245
Model Variables
The proposed model of the factors affecting cost and price escalation in higher
education contains a total of 42 variables, which are detailed below.
Applicants - Total number of students applying to an institution. This is influenced by Institutional Selectivity and the Per Capita Non-Tuition Subsidy.
Avg Peer Group Percapita Subsidy - The average per capita general subsidy provided by other members of an institution's peer group. This influences the assessment of the an Institution's tier quality ranking, as well as the institution's own Per capita non-tuition general subsidy.
Class Rank - The average high school class rank of those students accepted by the institution.
Current Gift Increases - Annual changes in current gifts resulting in the amount of current gifts donated to an institution for the current operating year.
Current Gifts - Annual gifts donated to an institution to be used for current operations. This changes annually based on increases or decreased donations through Current Gift Increases.
Endowment Increases Current Operations - Annual changes in donations for current operations from an institution's endowment. This impacts Total Endowment Support for Current Operations.
FTE UG Students - Total number of matriculated full and part time undergraduate students converted to a full time equivalent number.
Fed Govt Support Increase - Changes in Federal Government support for current operations. This impacts Federal Governrnent Appropriations. (Fed Govt Approp).
State Govt Support Increase - Changes in state government support for current operations. This impacts State Government Appropriations (State Govt Approp.)
Graduation Rates - The institution's five or six year graduation rate. It is affected by student quality indices of students matriculating into the institution (Student Quality). This impacts the institution's quality ranking (Tier Institutional Quality).
Higher Education and the Cost Disease 246
(1 1) Institutional Non-Endowed Aid - Amount of institution's current operating budget dedicated to student financial aid, not funded through gift sources of aid.
(12) Institutional Selectivity - The institution's acceptance rate. This impacts the Tier Institutional Quality ranking associated with an institution, as well as indices of student quality of students whom the institution accepts among its applicant pool. In turn, this is impacted and partially defined by the Per Capita Non-Tuition General Subsidy provided by the institution.
(13) Instructional Educational and General Costs - Total amount spent on Undergraduate Educational and General costs. This is the total amount spent on Personnel (Personnel Costs), non-personnel related expenditures (Non- Personnel Costs) and financial aid funded through the institution's operating budget (Institutional Non-Endowed Aid).
(14) "Non-Personnel Cost Increase Rate - The rate of increases in the cost of non- personnel related items purchased by the institution.
(15) Non-Personnel Costs - The amount of money spend on Non-personnel related items.
(16) PerCapita Peer Group I&G Expenditures - The Per Capita amount spent on Instructional and General Educational Expenditures among the peer group associated with an institution. The amount spent by peers impacts the amount of spending by the institution, since indices of quality are based on relative assessments among a peer group.
(17) Per Capita Educational and General Costs - The Per Capita amount spent on Instructional and General Educational Expenditures by the institution. This is derived by taking the total amount spent on Instructional and General Educational Costs and dividing it by the number of full-time equivalent students.
(18) PerCapita Inst Individual Subsidy - The average Per Capita amount awarded to students as individual merit or need-based aid. It is derived bv totaling the - amount of Non-Tuition revenue devoted to individual student aid (Total Non- Tuition Revenue not offered as a Non-Tuition General Subsidy, fiom Government Appropriations, Endowments devoted to current operations, Current Gifts, and the differences between the Tuition Sticker price and Net Tuition, based on a Full-time equivalency basis.
(19) PerCapita Net Tuition - The average net tuition undergraduate students actually pay the institution, formatted on a fill-time equivalency basis.
Higher Education and the Cost Disease 247
(20) PerCapita Non-Tuition General Subsidy - The average Per Capita amount granted to all students as part of a general subsidy provided by the institution. It is derived by totaling the amount of Non-Tuition revenue devoted the institutional general subsidy (from Total Non-Tuition Revenue, Government Appropriations and Current Gifts), based on a Full-time equivalency basis.
(21) PerCapita Revenues - Total revenue to support undergraduate education, on a 111-time equivalency basis. Derived by totaling all Non-Tuition and Net Tuition Revenues, divided by the number of full-time equivalent students.
(22) Personnel Cost Increase Rate - The average annual increase in costs for Personnel-related expenditures. Impacts the annual amount spent on personnel related expenditures.
(23) Personnel Costs - The annual amount spent on Personnel related costs.
(24) SAT Scores - The average SAT Scores of applicants accepted for admittance by the institution.
(25) EFR Student Quality - The average SAT and Class Rank of those applicants accepted for admittance. These indices are determined by the institution's standards of selectivity (Institutional Selectivity).
(26) Tier Institutional Quality - The Tier and Quality Ranking associated with an institution. This is impacted by Institutional Selectivity, the Per Capita Non- Tuition General Subsidy, as well as indices of Student Quality of the student applicants whom the institution accepts, and enrolls. The graduation rate M e r impacts measures of institutional quality. In addition, the Average Per-Capita Subsidy and Per Capita Instructional and General Educational expenditures provided by institutions among its peer group also help determine the relative quality ranking associated with an institution.
(27) Total Endowment Support Current Operations - The amount of endowment income devoted to current operations supporting undergraduate education.
(28) Local Govt Approp - Local Government Appropriations. The Total amount of local government appropriations supporting current undergraduate education.
(29) State Govt Approp - State Government Appropriations. The Total amount of state government appropriations supporting current undergraduate education.
(30) Total Govt Appropriations -The Total Amount of Government Appropriations devoted to supporting current undergraduate education. (Combined State and Federal Government Support).
Higher Education and the Cost Disease 248
(3 1) Total Non-Tuition Revenue - The total amount of revenue from non-tuition sources used to support undergraduate education. This is derived by totaling the amounts provided through Current Gifts, Total Endowment Support for Current Operations, and Total Government Appropriations.
(32) Total Percapita Subsidy - The total average amount of revenue used to subsidize undergraduate education, on a full-time equivalent basis. This is derived by adding the Per Capita Inst Individual Subsidy and Per Capita Non- Tuition General Subsidy.
(33) Total Revenues -Total revenue to support undergraduate education. Derived by totaling all Non-Tuition and Net Tuition Revenues.
(34) Total Tuition Revenue - The total amount of revenue received from tuition. This is derived by multiplying the Tuition Sticker Price by the number of full- time equivalent students.
(35) Tuition Based Indiv Aid - The average individual aid award offered to students from tuition derived revenue. This is derived based on the difference between the Tuition Sticker Price and Per Capita Net Tuition.
(36) Tuition Increase - This is the annual increase in the Tuition Sticker Price. Derived by subtracting the new year's Per Capita Educational and General Costs from the Total Per Capita Subsidy available to fund current operations.
(37) Tuition Sticker Price - An institution's published tuition and fee schedule for a particular academic year. In the model, this is derived by comparing the Current Tuition Sticker Price, the Total Per Capita Subsidy and anticipated Per Capita Educational and General Costs for the upcoming academic year. Any Tuition Increase is added to ensure revenues and expenditures balance.
(38) Per Capita Federal Student Aid - Average Per Capita award for Federal Student Aid.
(39) No. Fed Aid Recipients -Number of students awarded Federal student aid.
(40) Per Capita State Student Aid -Average per capita award for state student aid.
(41) No. State Aid Recipients -Number of students awarded state student aid.